HK1183548A - Computer-implemented land planning system and method - Google Patents

Description

Computer-implemented land planning system and method

The present application is a divisional application of an invention patent application having an application number of 200780008773.2, an application date of 2007, 30/1, entitled "computer-implemented land planning system and method".

Technical Field

The present invention relates to computer-implemented land planning systems and methods, such as those designed to generate at least one conceptual solution to a user-defined land development problem. The present invention is equally applicable to planning and developing single and multi-piece commercial, mixed use and residential land sites.

Background

Today the methods used by professional real estate developers, companies, government entities and other entities to assess the engineering feasibility, development costs and investment goals of the land are time consuming, inaccurate and expensive. Unfortunately, the current methods become even more complex and expensive due to bureaucratic political complications coupled with land use zoning, environmental protection requirements, extended approval procedures, and availability and escalating costs of land in the desired areas. This problem affects a wide range of land users, including, for example, real estate developers (office/industrial, commercial, retail, residential), companies and government entities (federal, state, county, city) that own and use real estate (public/private).

For each of the users described above, assessing the feasibility of a land site for development typically involves a land development team comprising one or more architects, engineers and land planners. Many of these team members participate in designing and planning the intended use of the site under consideration. This initial planning process may take 2 days to 4 weeks, typically producing a single schematic sketch with limited information (e.g., is a site supporting building footpoint (footprint) or building base (lot) and the necessary streets and/or parking bases. At this point, the developer will choose to contract additional plans and projects to more accurately assess the planning and budgeting feasibility, based largely on intuition and "feeling of mind" on the project. This process may take 2 to 16 weeks and usually results in only one option, which is based on the experience of the designer but is not optimized in any way. This information is then used to estimate a more accurate budget. Often requiring time value engineering to bring the design back within the original budget. This process takes 2 to 6 weeks. The final budget is generally not determined until the planning process is complete, sometimes 3-4 months after the land site was initially considered.

The above planning process often must occur prior to purchasing ownership and requires significant investment and security to maintain ownership for an extended period of time at a legal cost.

After these 4 to 28 week processes (16 weeks on average) and considerable expense and risk of losing opportunities, developers must evaluate the risk of purchasing and developing ownership based on a design option that is not optimal. Unfortunately, the process outlined above is complex and even more responsible due to miscommunication and inability to communicate among many involved groups, which often leads to bad designs, bad budgets, misopinions, and bad projects.

The applicant has realised that the land development industry requires major paradigm changes and it is now possible to do this by improvements in mathematical modelling and computational hardware. It is a primary object of the present invention to identify the problems outlined above by means of a virtual engineering system that can generate many optimization alternatives for land development, including planning, engineering and budgeting for each potential solution. This calculation is typically carried out over a period of up to 24 hours.

Heuristic strategy

The speed and effectiveness of the present invention is advanced by utilizing heuristic mathematical optimization methods such as, for example, evolutionary algorithms (with possible instantiations such as, for example, genetic algorithms, evolutionary strategies, evolutionary programming, genetic programming, and combinations and components thereof). For certain subtasks, mathematical programming methods such as, for example, linear programming, mixed integer programming, branch-and-bound programming are also used.

Briefly, evolutionary algorithms (or "EAs") are planning techniques that mimic biological evolution as a problem solving strategy. Given the specific problem that needs to be solved, the input to the EA is a set of potential solutions to that problem encoded in some form, and a metric called the fitness function (fitness function) that quantitatively evaluates each candidate. These candidates may be known to be working solutions, the purpose of the EA being to improve them, but more often they are generated randomly.

From these initial candidate solutions, replication is performed in such a way that better candidate solutions receive on average more duplicates (according to their fitness measure) and poorer candidate solutions receive on average less duplicates, through a process called replication. Furthermore, rendering may not be based on fitness (fitness), but rather may choose a solution entirely randomly from the parent population. These copies, produced by the rendering, enter the next generation of the algorithm and then undergo a modification process of randomization called mutation and crossover (also called recombination). Once again, the newly created solutions are quantitatively evaluated to determine their fitness values, at the mutation and crossover (often referred to as "change operators"). After this step of fitness determination, a selection step may be added, either to determine or according to a process of fitness-based randomization-selecting a better solution from the offspring population to live, while discarding the poorer solution. This selection step may be applied to the descendants only, or to a union of parents and descendants. This process is then repeated. It is desirable to increase the average fitness of the population per round, so by repeating this process hundreds or thousands of rounds, a very good solution to the problem can be found.

Disclosure of Invention

It is therefore an object of the present invention to provide a computer-implemented land planning system and method in which at least one solution conceptually appropriate to a user-defined land development problem may be automatically generated in one exemplary embodiment.

It is a further object of the present invention to provide a computer-implemented land planning system that, in one exemplary embodiment, may utilize heuristic problem-solving strategies such as, for example, evolutionary algorithms. According to one evolutionary algorithm, evolution starts with a completely random population of individuals and occurs in generations. In each generation, fitness of the entire population is evaluated, and multiple individuals are randomly or deterministically selected from the current population (based on their fitness), modified (mutated and/or recombined) to form a new population, either in whole or in part, which becomes the trend for the next iteration of the algorithm. During this process, the size of the population may remain unchanged (as in genetic algorithms) or may vary (as in (μ, λ) -or (μ + λ) -evolution strategies).

It is a further object of the present invention to provide a computer-implemented land planning system in which land planning and engineering may be performed simultaneously in one exemplary embodiment. This invention can consider various land development parameters (e.g., site specifications, user constraints, cost information) in advance from the perspective of land planners and engineers, and thus explore thousands of options using heuristic algorithms to determine which options are the best options determined in terms of cost and/or revenue.

It is another object of the present invention to provide a computer-implemented land planning system that in an exemplary embodiment can apply heuristic problem-solving strategies to current civil engineering processes to make fundamental changes to residential and commercial land planning and development.

It is another object of the present invention to provide a computer-implemented land planning system that in one exemplary embodiment can reduce the time it takes to obtain a final project draft (85% or more complete) containing cost information, in many cases from 3-4 months to less than 24 hours.

It is a further object of the present invention to provide a computer-implemented land planning system which, in one exemplary embodiment, can provide web-accessible technology that can ensure that a user determines the most cost-effective way to develop a land site.

It is another object of the present invention to provide a computer-implemented land planning system that, in one exemplary embodiment, can visualize land development issues and propose a final solution.

It is another object of the present invention to provide a computer-implemented land planning system that, in one exemplary embodiment, can provide land developers with direct access to qualified information in approximately 24 hours (or less) rather than in multiple months.

It is another object of the present invention to provide a computer-implemented land planning system that, in one exemplary embodiment, minimizes the initial investment capital required to develop a land site.

It is a further object of the present invention to provide a computer-implemented land planning system that, in one exemplary embodiment, can reduce construction costs.

It is another object of the present invention to provide a computer-implemented land planning system that, in one exemplary embodiment, minimizes the risks associated with developing a land site.

It is a further object of the present invention to provide a computer-implemented land planning system that, in one exemplary embodiment, can minimize project time.

It is a further object of the present invention to provide a computer-implemented land planning system in which the creative (aesthetic) and engineering aspects of land planning and development can be efficiently integrated in one exemplary embodiment to achieve a very good or even globally optimal solution.

It is a further object of the present invention to provide a computer-implemented land planning system in which financial measures such as, for example, cost and/or Return On Investment (ROI) may be optimized in one exemplary embodiment.

It is another object of the present invention to provide a computer-implemented land planning system that in one exemplary embodiment can generate multiple "optimally different" solutions to the land development problem and provide these solutions in a ". dwg" format that can be input to and directly processed by an engineer's existing CAD/CAM system.

It is a further object of the present invention to provide a computer-implemented land planning system that, in one exemplary embodiment, may be used on a stand-alone PC or may be used on a network.

It is another object of the present invention to provide a computer-implemented land planning system that, in one exemplary embodiment, can utilize digital satellite topography.

It is another object of the present invention to provide a computer-implemented land planning system that can utilize, in one exemplary embodiment, heuristic problem-solving strategies that are capable of processing many parameters simultaneously.

It is another object of the present invention to provide a computer-implemented land planning system that, in one exemplary embodiment, can utilize heuristic problem-solving strategies that search on top of local optima.

It is another object of the present invention to provide a computer-implemented land planning system that in one exemplary embodiment can utilize a heuristic problem-solving strategy designed to find a global optimum-an attribute referred to as global convergence with probability 1-within a space with many local optima.

It is another object of the present invention to provide a computer-implemented land planning system that, in one exemplary embodiment, can utilize heuristic problem-solving strategies applicable to traffic engineering including signal optimization and highway design.

It is another object of the present invention to provide a computer-implemented land planning system that, in one exemplary embodiment, can utilize heuristic problem-solving strategies that can be used to optimize the design of buildings and bridge structures.

These and other objects of the present invention are achieved in the exemplary embodiments disclosed below by providing a computer-implemented land planning system that is designed to generate at least one solution that conceptually fits a user-defined land development problem. In one embodiment, the system uses a computer readable medium and a computer program encoded on the medium. The computer program is operable when executed on a computer to electronically create at least one candidate solution to a land development problem. The candidate solution mixes a plurality of engineering measures applicable in the development of the undeveloped land site. The fitness function quantitatively evaluates the candidate solution based on its cost. These fitness functions may also contain one or more penalty components that state that the candidate solution violates one or more user-defined constraints. A heuristic problem-solving strategy processes the engineering measurements of the candidate solutions to obtain a more quantitative solution that fits the land development problem. Output devices such as display monitors, printers, electronic communications, etc. communicate files to the user illustrating solutions appropriate to the land development problem.

The term "planning" is broadly defined herein to indicate the development of any conceptual land site. The term "undeveloped land site" refers to a site that may or may not have an existing structural and/or engineering architecture and that has not been developed with a suitable solution generated in accordance with the conceptual present system. The term "heuristic" broadly refers to any problem solving strategy that utilizes adaptive, self-learning, or adaptive techniques (such as feedback evolution) to improve performance. The following are examples of heuristic problem-solving strategies: evolutionary algorithms (such as, for example, genetic algorithms, evolutionary strategies, evolutionary programming, genetic programming, and variants of these), simulated annealing, differential evolution, neural networks, hill climbing strategies, and ant population optimization, particle swarm optimization, and tabu search. For some subtasks, linear programming, mixed integer programming, and branch-and-bound algorithms are also considered heuristic algorithms.

According to another exemplary embodiment, a tool, such as, for example, a digital terrain model, digitally represents an undeveloped land site in three-dimensional space.

According to another exemplary embodiment, the computer program comprises instructions for conceptually positioning the engineering measurements within a three-dimensional space.

According to another exemplary embodiment, the engineering measurements are selected from the group including, but not limited to, storm water systems, sanitary sewage collection systems, and potable water systems.

According to another exemplary embodiment, the output file includes at least one computer-generated draft.

According to another exemplary embodiment, the output file further comprises a list of costs listed itemized in the engineering measurements.

According to another exemplary embodiment, the file is delivered to the user over a global communications network.

In another embodiment, the present invention is a computer-implemented land planning system designed to generate at least one conceptual solution to a user-defined land development problem. The processor obtains land development constraints for the undeveloped land site. The system further uses a computer readable medium and a computer program encoded on the medium. The computer program is operable, when executed on a computer, to create a population of candidate solutions to a land development problem. Each candidate solution contains a plurality of engineering measurements applicable to the development of an undeveloped land site. The processor obtains a cost model containing individual cost data for each engineering measurement. The computer program includes inappropriate solutions for penalizing (or even discarding) violations of land development constraints. For all solutions that have not been discarded immediately due to violations of constraints, a fitness function is used to calculate a fitness score based on engineering measured cost data. The fitness function computes the fitness of the candidate solution using various cost measures and also using various penalty measures. A heuristic problem-solving strategy processes the engineering measurements of each selected candidate to obtain an increased fitness score, so that those candidate comparisons that achieve the increased fitness score are likely to be used or even deterministically selected to form a new population in the next iteration of the algorithm. The computer program comprises means for selecting a set of optimally different alternative solutions from a plurality of suitable solutions. Output devices such as display monitors, printers, electronic communications, etc. are used to document the user with the best different alternative solutions illustrating the land development problem.

According to another exemplary embodiment, the processor obtains user preferences for an undeveloped land site.

According to another exemplary embodiment, the computer program includes instructions for penalizing the fitness score of the candidate solution based on violating user preferences.

In yet another embodiment, the present invention is a computer-implemented land planning method designed to generate at least one conceptual solution to a user-defined land development problem. The method comprises the step of electronically creating at least one candidate solution to the land development problem. These candidate solutions include a plurality of developed engineering measurements that can be used at undeveloped land sites. The candidate solutions are evaluated quantitatively based on their overall fitness (which is composed of a cost component and a penalty function component in turn). The engineering measurements of the candidate solution are then processed using heuristic problem-solving strategies to obtain a more quantitative solution that fits the land development problem. After the more appropriate solution is obtained, a file illustrating a solution appropriate to the land development problem is output to the user.

Brief glossary

The following glossary provides the basic definitions/explanations of some of the terms included in this specification:

cellular automata-typically a group of algorithms on a grid that perform some global task using a set of "cells", each of which has the same local performance. They are often used to mimic natural forces such as, for example, the attractive forces on sand or water droplets in a rain cloud.

Cross-according to evolutionary algorithms, this refers to the process of combining multiple parts of two or more individuals (in this case parents) to produce one or more new individuals (offspring).

Delauney triangulation-a method of generating triangles in a unique and mathematical way in two dimensions with line junctions.

Amenities-areas of land with special ownership that can be considered optimized. Often work rights include, for example, tree protection areas, power transmission lines, existing natural gas and water pipelines, and natural elements like creeks.

Evolutionary Algorithm (EA) -a group of algorithms that use the concept of natural evolution in an abstract way to optimize solutions to often complex problems. As discussed below with reference to fig. 22-25. Subsets of evolutionary algorithms include, for example, evolutionary strategies, genetic algorithms, genetic programming, and simulated annealing.

Evolution cycle-an iterative cycle used according to an evolution algorithm.

Evolutionary strategies-a subset of evolutionary algorithms that can use a continuous search space, generally rely more on variation than crossover, and often use an appropriate strategy of step size to evolve parameters and increase performance.

Fitness (function) -fitness of an individual defines the level of the individual in a population and has a probability that the individual is selected for pairing. The fitness function is a function that calculates the fitness of an individual.

Heuristics-heuristics are implementations of assumptions that help the optimization process. The Layout solution process (Layout Solver) may, for example, use the assumption that minimizing sidewalks always minimizes costs.

Individual-according to evolutionary algorithms, this is the entity that evolves. It may be constituted by input parameters for the fitness function. The newly created individuals are generally referred to as "offspring" and the individuals selected for pairing are generally referred to as "parents".

Iterative loop-a loop that makes a series of possible options or solutions.

Multiple-block business-a multiple-block business location may be a location where multiple commercial buildings are developed. These buildings may be connected. Also, a building having different floors as separate parts of a single building is often referred to as "multi-block".

Variation-small variation in individuals. What often happens after individuals are generated with crossover or ordinary replication.

Niches-methods that not only optimize the best solution in a population, but also preserve the diversity of the population. This results in a so-called "optimally different" solution, a group of different solutions that are all locally optimal.

Best-different-the best-different solution is the solution that is "best" in the neighborhood of its search space. That means that a set of optimally different solutions shows not only the best solution in the whole search space but also an extension of the locally optimal solution.

Optimization-a process that attempts to find the optimal solution in a search problem. In one embodiment, the optimization process is an attempt to find the optimal layout, level and utility plan that optimizes land development costs. The cost calculation is performed using simulation.

Simulation-applying mathematical rules to a process that mimics the behavior of a real-world process. In this case, the simulation may be tuned towards "mimicking" the structure of the site and calculating the cost of this structure.

Monolithic business-a monolithic business site is a site where only one building is developed.

Detailed Description

The present invention now will be described more fully hereinafter with reference to the accompanying drawings, in which one or more exemplary embodiments of the invention are shown. This invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein; rather, these embodiments are provided so that this disclosure will be thorough and complete. Like numbers refer to like elements throughout. As used herein, the article "a" is meant to encompass one or more items. Where only one item is intended, the use of "a" or similar language. Although specific terms are employed herein, they are used in a generic and descriptive sense only and not for purposes of limitation. Unless otherwise specifically defined herein, such terms are intended to be given their ordinary and customary meaning as is not inconsistent with the application to the relevant industry and are not limiting of any specific embodiment described herein. Any reference to an advantage, a benefit, an undesirable result, or the operability of the invention does not imply a statement that the invention has previously become practical or that any testing has been performed.

In an exemplary embodiment, the present system operates in an environment utilizing client devices that communicate with a host server over a computer network such as, for example, the Internet. In other embodiments, other computer networks may be used, such as a Wide Area Network (WAN), a Local Area Network (LAN), or an intranet. The host server may include a processor and a computer readable medium such as, for example, Random Access Memory (RAM). The processor is operable to execute certain heuristic problem solving programs and other computer program instructions stored in the memory. Such a processor may include a microprocessor (or any other processor) and may likewise comprise, for example, a display device, a data storage device, a cursor control device, and/or any combination of these elements, or any number of different elements, peripherals, and other devices. Such a processor may also communicate with other computer-readable media storing computer program instructions, such that when the stored instructions are executed by the processor, the processor performs the steps described herein. Those skilled in the art will also recognize that the exemplary environment described herein is not meant to limit the application of the present system, and that alternative environments may be used without departing from the scope of the present invention.

Various problem solving programs incorporated into the present system and discussed herein utilize data from a data storage device as input. In one embodiment, the data storage device comprises an electronic database. In other embodiments, the data storage device may comprise an electronic file, a magnetic disk, or other data storage device. The data storage device may store engineering and cost modules, building codes and rules, user data and libraries. The data storage device may also contain other items for performing the functions of the present system.

In one example, the problem-solving program includes one or more heuristic problem-solving strategies (particularly, evolutionary algorithms such as, for example, evolutionary strategies, genetic algorithms, evolutionary plans, genetic plans, and heuristics) to solve high-level problem statements defined by the user-e.g., optimizing land development of the site based on cost. The generated optimally different solutions are transmitted to the client device over the computer network. The user can then decide which of the suitable solutions best meets his or her design objectives.

I. System overview

The present system uses an optimization engine that is divided into three different solutions processes discussed separately below. These solutions allow the user the option of implementing only a desired portion of the overall functionality of the system and thereby speed up the optimization process.

Three separate solutions include:

(a) planning a solution process

(b) Grading solving process (grading solution)

(c) And utility solution process

Planning a solution process

Planning a solution process to plan a place; parking spaces, sidewalks, driveways, highways, and other 2D features are added to the venue. The aim is to optimize the position of the buildings of the site, given all the planning constraints entered by the user.

Hierarchical resolution process

The hierarchical resolution process optimizes the levels proposed by the site given a certain fixed plan, making the earthwork feasible and optimal. This solution takes into account user constraints such as minimum and maximum slopes, retaining walls and fences, for example.

Utility solution process

The utility solution process optimizes the pipes and inlets to the site. This solution takes into account factors including pipe size, depth, water flow on the surface and flow in the pipe.

A.Different applications

The three solutions may be used not only alone, but also in a stack combining two or all three to obtain more detailed information. Stacking the solutions affects the complexity of the optimization and therefore the speed of computing one or more "good solutions".

For the added complexity, the resolution process can be used online and offline:

online: the resolution process begins within the browser window. In this embodiment, during runtime, a user may interact with the solution process and may visualize the evolution or computation of different solutions.

Off-line: the resolution process begins by sending a request to the server farm. In this embodiment, the results of the optimization are sent back to the user when the request has been completed.

Solution processes are not typically stacked when online, as stacking solution processes adds complexity to the optimization. Running online may be considered a "quick resolution process" or "simulation". FIG. 1 illustrates user access to online and offline optimization/simulation. A detailed explanation of the resolution process stack is provided elsewhere herein.

To promote portability and due to the relative complexity of the computation, the system may utilize a global communication network such as the internet, for example. To use the system in an embodiment, a user uses an internet browser and makes internet access. The user may create a project and store data in the host database using the web interface. With a web interface, a user can quickly run a completion resolution process in less than 5 minutes. These quick resolution processes provide advanced information to the user and can be used to address basic feasibility and aesthetic considerations. If more detailed information is desired, the user may request an offline solution from the stacked solutions process. Stacked solution process solutions provide engineers with more details, cost, and options to use in their site planning tasks. Typical runtime can range from 10 minutes to 24 hours depending on which (stacked) resolution processes are used in such a run and the size and complexity of the submitted site.

B.Stack resolution process

As mentioned above, there are different ways to stack up the three solutions in the present system. Fig. 2 shows a general interface for the different solutions.

There are 6 different scenarios that can be used:

(1) using only planning solutions

(2) Using only hierarchical resolution processes

(3) Using utility only resolution processes

(4) Stacking planned resolution processes on a hierarchical resolution process

(5) Stacking hierarchical resolution processes on utility resolution processes

(6) Complete heap using planning solution process, hierarchical solution process, and utility solution process

Each of these operational scenarios requires different inputs and generates different outputs. Generally, the input relates to the first solution process in the heap, and thus the result is a combination of the results of all solution processes in the heap. If two resolution processes are stacked, the first one uses the second as a "fitness function". This concept will be discussed further below.

C.Planning only the solution process

Planning solutions by self-running is often done to quickly assess the general aesthetics or feasibility of certain site characteristics. Only basic information for a given site is entered, such as, for example, ownership boundaries and building, contours, planning solution process operations to determine the best location for a building, taking into account the minimum number of parking spaces, motorways, truck return lanes (if needed), floor power, and other user-defined two-dimensional constraints.

Inputting:

-ownership boundaries

-building outline

Two-dimensional constraints which can be selected (e.g. work area, motorway requirements, parking space requirements)

And (3) outputting:

-finding an optimal plan based on the number of parking spaces and the paving area.

When this resolution process is run as a fast resolution process in the web interface, the user also has the option to interact with the resolution process and move the building. The planning solution process updates the plan quickly as the building is moved, with each new location selected by the user. In this embodiment, the user can quickly assess the feasibility of the site and the general location of the building. Since any change in two-dimensional constraints like e.g. the number of parking spaces or the motorway requirements also triggers an update of the plan, this interface gives the user a powerful tool to answer a number of feasibility questions, including e.g.:

can 100 parking spaces be equipped at a site with building option a?

Can 100 parking spaces be equipped at a site with building option B?

Can 100 parking spaces be provided at a site with building a, two motorways and provide truck access behind the building?

Can 100 parking spaces be provided, two motor lanes be provided, truck access be provided, and a drain pool be installed at a site having building a?

Fig. 4 and 5 provide examples of the previous and subsequent plans, respectively. Before the planning solution process runs (fig. 4), the user must enter only basic characteristics of the site overall, such as data like ownership boundaries, buildings, and topography. After running the planning solution process (fig. 5), the system has created a stop line, parking area, vehicle access, truck lane to building, and sidewalk. The user can control the options of the planning considered by providing the system with rules to follow, such as:

the building moves back 150 feet from the line of ownership.

No parking is allowed behind the building.

The truck access must be 60 feet wide.

D.Only hierarchical resolution processes

The self-running hierarchical resolution process gives a quick indication of the planned land segmentation and land fill costs provided. In this scenario, the user enters a grading constraint such as, for example, a minimum value for a parking lot slope, a maximum value for retaining wall height, and so forth. When doing this optimization, the system considers the overall grading of the lifting/lowering ground, which places retaining walls, which needs to be on the curb, the appropriate slope for parking, and other related grading issues. The hierarchical resolution process then considers planning the different areas specified, reading the user-defined hierarchical constraints, and cost-optimizing the hierarchical planning of the site.

Inputting:

-completion of the planning

-three-dimensional existing grading

-hierarchical constraints of the planning

Individual unit cost of grading (digging/filling) a site

And (3) outputting:

-proposed three-dimensional grading

Cost reports on proposed rankings (except utility cost)

Running this quick resolution process through a web interface for users provides a three-dimensional hierarchical browser that allows them to view actively ranked places. Fig. 6 and 7 illustrate the image demonstration operations of the 3D grading simulation tool before and after. This tool visually illustrates the land being divided/filled, any excavated rock, where the retaining wall is placed, where the curb is built, where the motor vehicle is driven, and other details.

The rating options considered by the system may be customized by the user. The user can type the rules into a system that directly rates the venue. Such rules include, for example:

-specifying a maximum retaining wall height

Minimum and maximum values of the allowed slope indicating the parking area

To further optimize the staging plan, a staging solution process and a utility solution process may be run in a stacked scenario such that the staging takes into account the drainage problem in its optimized solution.

E.Utility-only solution process

The utility solution process generates an optimized drainage plan including inlets, outlets, required to efficiently drain the site. Utility solutions processes stacked below the hierarchical solution process are typically run. The utility solution process may also run as a quick solution process on the web interface by itself to provide valuable feedback on aesthetic and feasibility issues.

Inputting:

substantially complete planning

Fully complete proposed three-dimensional grading

-utility constraints of the plan

Cost per unit of utility design (piping/inlets, etc.)

And (3) outputting:

-three-dimensional utility plan

Cost reporting of utility plans

Fig. 8 and 9 provide examples before and after the options from the utility resolution process, respectively. Figure 8 shows a planned site without any utility design provided. FIG. 9 shows the site after the utility resolution process has completed the design. Fig. 10 shows details provided to the user in the figure.

A user may specify a certain utility design by entering constraints into the system. Some constraints that may be established include:

minimum and maximum values of the pipeline slope

Cost per diameter of pipe

F.Planning + grading solution process

Stacking the planning solution process on top of the hierarchical solution process requires the system to rank each generated feasible plan. This allows the user to consider the lowest cost, feasible plan according to the hierarchy. This operational scenario is generally preferred for measuring only the value of the parking space or the size of the road paving area. It is also ensured that the place with the feasible rating is selected as the best option. Since a large number of planning/hierarchy combinations are run for this scenario, the stacked solution process returns not only the best plan, but also 5 to 10 "optimally different" proposed solutions. In this case, the stacked resolution process returns a different planning alternative with comparable costs, giving the user more insight into options regarding resolution of civil engineering problems.

Inputting:

-ownership boundaries

-building outline

-three-dimensional existing grading

Selectable two-dimensional constraints (e.g. ground-work area, motor vehicle requirements, parking space requirements)

-constraints of the planning hierarchy

Individual unit cost for grading (digging/filling) sites

And (3) outputting:

5 to 10 optimally different solutions. Fig. 11 and 14 illustrate two "optimally different" solutions.

Output of solutions (for each solution the following options may be provided):

complete site planning with buildings, ownership borders, parking, motor vehicles, work-works, ponds, etc

-complete three-dimensional proposal ranking

Full HTML and Excel cost report which summarizes all expenses incurred in rating and planning the proposed site

Full DWG files with hierarchies defined for each planning element, hierarchical outline, text detail required for civil engineering

A full system file containing all relevant details about the site that can be fed back into the system for further refinement

Block diagrams of detailed solution outputs are provided in fig. 11-16. This stacked combination is run off-line and typically within 1 hour.

G.Grading + effect solving process

Combining the grading resolution process with the utility resolution process requires that the grading resolution process take into account the impact of grading on the drainage plan. If the grading is changed, the water may flow in a different direction or area in the field. Even though the grading may be inexpensive, this may require a more expensive drainage scheme. By running utility solutions for different tiers during the tier optimization, the utility solutions process can highlight problem areas in the tiers that may be expensive to solve. Those regions are then ranked differently by a ranking resolution process to minimize overall cost.

Inputting:

complete planning

-existing grading in three dimensions

-constraints of the planning hierarchy

Individual unit cost for grading (digging/filling) sites

-utility constraints of the plan

Cost per unit of utility design (e.g. piping and inlets)

And (3) outputting:

an optimized option for the provided site planning

Output of solutions (for each solution the following options may be provided):

full site planning with buildings, ownership borders, parking, motor vehicles, work-works, ponds, etc

-ranking of full three-dimensional proposals

-a fully three-dimensional proposed utility plan

Complete HTML and Excel cost reports, which outline the grading and all the expenses incurred for planning the pipelaying of a proposed site

DWG document defining a hierarchical completeness of the hierarchy for each plan element, hierarchical outline, piping diagram and document details of all the areas required for civil engineering

A full system file containing all relevant details about the site that can be fed back into the system for further refinement

FIGS. 11-16 provide detailed block diagrams of the solution outputs. This stack solution typically runs within 30 minutes.

H.Planning + grading + utility solution process

With a complete solution process stack it is possible to obtain 5 to 10 completely optimized plans with minimal information. In this scenario, the grading resolution process also takes into account the cost of the drainage plan. This then completes a complete optimization of the feasible site plan that provides the lowest cost, taking into account all site planning, ranking, and utility constraints/costs. In general, this full stack is more complex and requires more time to compute, and will provide the most accurate and detailed results.

Inputting:

-ownership boundaries

-building outline

-three-dimensional existing grading

Selectable two-dimensional constraints (e.g. ground-work area, motor vehicle requirements, parking space requirements)

-hierarchical constraints of the planning

-utility constraints of the plan

Unit cost for grading (digging/filling) sites

Cost per unit of utility design (e.g. piping and inlets)

And (3) outputting:

5 to 10 optimally different solutions. Fig. 11 and 14 illustrate two "optimally different" solutions.

Solution outputs (for each solution the following options may be provided):

complete site planning with buildings, ownership borders, parking, motor vehicles, work-works, ponds, etc

-ranking of full three-dimensional proposals

-a fully three-dimensional proposed utility plan

Full HTML and Excel cost reports, which summarize all expenses for site planning proposed by staging and pipelaying

Full DWG document with hierarchical definition of the text details of all the areas required for each planning element, hierarchical outline, piping diagram, civil engineering

Full system files containing all relevant details about the site that can be fed back into the system for further refinement

Detailed block diagrams of the solution outputs are provided in fig. 11-16. Running a full pile for a 10-15 acre site may require between 8 and 15 hours.

Optimization of

Multiple optimizations occur in different solutions. The following discussion outlines how these optimizations can be made and settings that may be used.

(i) Evolution cycle

Evolutionary Algorithms (EAs) use evolutionary cycles to evolve a pool of individuals that utilize "selection", "replication", "recombination" and "mutation" in an effort to continually improve the individuals and find the optimal solution. The augmented introduction of an EA is provided separately below and thus is not repeated, but rather the following discussion focuses on how the planning solution process implements this evolutionary process.

Fig. 3 shows a general evolutionary cycle. First, an initial population, or "pool" of individuals, is generated. The size of this population is generally fixed throughout evolution. The quality or "fitness" of each individual is measured using a fitness function, after which termination criteria are detected. Typically, the cycle is terminated after several generations, but the fitness or change in fitness over time at the top may also play this role. Second, the "best" individual is selected. The optimal set of individuals does not always have to be the set of individuals with the best fitness (see discussion on "optimally different" and "niche" (which may not be the case here)). The selected individuals are used to form the next generation. This can be done with recombination, but replication is also often used. The next generation is typically the same size as the first generation, but is generated using only the best individuals of the first generation. This next generation will be mutated in the hope of changing in the right-hand (or better) direction. This way of individual variation may be a relevant problem, but the use of "step size" in most evolutionary strategies affects the size of the resulting variation. Strategically, changing the size of this step size can therefore significantly affect the performance of the algorithm. After mutation, this cycle is stopped by recalculating the fitness of each individual. Since this pool now consists of the latest generation of near-optimal volumes, it is expected that the fitness value of this generation will now be improved.

In one embodiment, this evolutionary process is used in areas where all other optimization processes fail to proceed. This means that if an alternative to the mathematical approach is found for a certain subset of problems, this alternative approach is generally preferred. The evolution process is typically not deterministic and will therefore not generate the same answer twice. And because the search space for most real-world questions is too large to consider all possible answers, a solution generated by an evolutionary algorithm may never prove to be the best solution. Although evolutionary algorithms are relatively flexible and perform well where other optimization algorithms are not available to find a solution.

(ii) Optimization and simulation

The present optimization engine is both an optimizer and a simulator. These two concepts are often confused and are therefore explained below.

Simulation according to a computer program is a process that mimics what would happen given certain inputs. This may be the case of analyzing an industrial process over time, computing a weather model or in such a case, attempting to apply "public feelings" to land development problems. Simulations are usually deterministic and give a unique answer.

The optimization process according to the computer program is a process of optimizing what can be given constraints. An important difference from simulation is the fact that the computer attempts to find the best input to match the problem and its constraints, rather than analyzing a given set of inputs.

The present planning solution process has both optimization and simulation elements, and in some parts the solution process can do a bit for both. How the solution process operates is based on certain assumptions made by the system. A constraint optimizer for each hypothesis to be valid; it tells the optimizer that he does not have to find the best solution. Each hypothesis thus reduces the search space and generally increases the speed of the optimization process. However, assumptions of how many choices cannot be made may have a significant impact on the quality of the candidate solution. Effectively, each assumption of manufacturing moves part of the problem from the optimization process domain to the simulation domain. In other words, rather than attempting to find the best solution to the sub-problem, it is assumed that it is optimal in a manner that makes the solution to the sub-problem deterministic and more simulation than optimal.

A.Planning a solution process

The optimization elements of the planning solution process will be described below. Fig. 17 summarizes the planning solution process and its internal workings.

The optimization element of the planning solution process is the mechanism that seeks to find the best location of the building. To measure the quality of the building location, the simulation element of the planning solution process is used to mimic where the rest of the planning will proceed as a result of the building location. According to the Evolutionary Algorithm (EA), the optimization element is an "evolutionary loop", while the simulation element comprises "calculation details" and considers "individuals" as specific locations of buildings at the site. The details and prescribed settings for this method are given in the following discussion.

(i) Individuals

The individual in the optimization process of the planning solution process is the location of the building. In the case of multiple commercial sites, this may be the location of multiple buildings. In the case of a business, the user may also wish to optimize the location of the pond or the advertising symbol so that these items become part of the individual.

The position of the building is defined as the center of the building from the northeast aspect, plus the rotation of the building for a given original shape.

(ii) Variation of

The individual mutation is performed by moving and rotating the building. The shift and rotation amounts are gaussian random distributions with a standard deviation of a certain step size σ. This step size starts with a reasonable size, e.g., the diagonal/λ of the venue, where λ is the population size, which can be varied according to some strategy. One successful strategy used is "adaptation," which uses the principle of evolution to evolve the optimal step size while evolving at the location of the building. Fig. 18 shows an example of the location of a building and its replacement.

The variability of individuals is constrained by many real-world logistic problems. Most of these problems are detected only by attempting to place vehicles, islands and stops on the premises; but some are done before the fitness estimation, like detecting if the building is within the building back-off. Infeasible locations, such as this, are not considered in the fitness estimation.

(iii) Optimally different choices or "niches"

In one embodiment, the selection in the planning solution is made according to an area known as a "niche". The planning solution process not only finds the solution that is the best to compute, but there are values in returning different solutions. For civil engineers or land developers, 5 different "very good" solutions may provide a better understanding of the possibilities for a given site than 5 very similar (perhaps better) solutions.

The optimal generation of different solutions is obtained by re-using the definition of "good individuals" formulated: the individual with the best fitness is in its local neighborhood. The term "niche" indicates this local neighborhood, and hence the term "niche". During selection, as opposed to selecting the μ (parent) individuals with the highest fitness, the selection is now made to select only the one best individual, as is done normally. All individuals in the immediate neighborhood of this best individual are now removed and the best remaining individual is selected. This process is repeated by removing the immediate neighborhood of the second individual and selecting again until all individuals are removed or μ individuals are selected.

Defining the neighborhood of an individual is an important step in this type of selection. In general, the neighborhood is defined by a radius ε. In the case of a northeast position of the building, this relative is straightforward, but in the case of rotation, it becomes more complex.

Consider two buildings in the same neighborhood if and only if:

-centres no more than epsilon (diagonal of the venue/c) feet apart

-the rotations differ by an angle not greater than 45 degrees

(setting c closely related to the number of parents selected; c = mu is not uncommon)

If the number of niches falls below μ, the best individual not selected in each niche is selected until μ is met.

(iv) Fitness function

The fitness function of the planning solution process simulates the effect of the location of the building on the location of the remaining features of the plan. The fitness may be measured differently depending on different constraints set in the building. The stacking of solutions also affects how the fitness function is defined. If the planning solution process is run on its own (by itself), no cost is calculated. If the planning solution process is stacked on top of the hierarchical solution process, the fitness is the minimum or optimal hierarchical cost. If the planning solution process is stacked on top of the hierarchical solution process and the utility solution process, the fitness is the minimum/optimal total cost.

The minimum number of parking spaces required by a building defined by the user changes the definition of the fitness of the stand-alone planning solution process. If this quantity is not defined, the solution looks for the maximum number of parking spaces using the entire facility. If this number is set, the solution process uses sub-areas of the site to find the minimum of the area required for the parking space. When the planning solution processes are not stacked, this is used as the fitness of the building location.

(v) Iterative loop

By eliminating some potential building locations using heuristics, the task of considering all the remaining building locations on the site becomes feasible.

A significant difference between this method and the evolution cycle is the way in which the building location is defined. The north, east and rotation are no longer performed, but by selecting two sides of ownership, one side of the building and two offsets.

It is generally more productive to utilize this iterative loop rather than the evolutionary loop. However, this method requires many assumptions in order to iterate over the search space. Another concern is that the now simplified problem may still have a very large number of possibilities, or candidate solutions. The complexity is reduced to polynomial O (n4), but its complexity can grow rapidly. As such, this method is particularly useful in single-block commercial land development.

In some cases, the user already knows which side of the building needs to be aligned with which edge of the ownership boundary. This input further reduces the search space. The complexity in this case is reduced to O (n 2). Since this results in a large reduction in the number of candidate solutions, the quality of these solutions can be improved by reducing the step size in the offset and increasing the offset.

A similar approach can be used to improve the quality of smaller sites. If the number of possibilities is small, the step in the offset can be reduced and thus the number of steps can be increased.

B.Hierarchical resolution process

Although the hierarchical resolution process can be used as part of the fitness function by the planning resolution process, the hierarchical resolution process can also be considered as its own optimization process. It optimizes the earthwork of a site given certain slope and height constraints. The optimization process of the hierarchical resolution process will be discussed below.

A rating is defined as a "surface" and all of its retaining walls. To optimize the hierarchy, the following assumptions are made about the data structure:

the retaining wall being located only around the two-dimensional area

The surface being formed by a triangular mesh

Each point (except those on the retaining wall) has a unique surface height

The retaining wall is typically located around the parking lot and any exceptions can be handled by segmenting the area where the retaining wall needs to be located. The use of triangular meshes to define surfaces is common in civil engineering and is often referred to as TIN.

Because the retaining wall is defined to lie on the boundary of a two-dimensional area, the retaining wall between areas a and B may be defined by two rows of three-dimensional points; one row in region a and one row in region B, both having the same two-dimensional position but different heights. The difference in height creates vertical walls. This ensures that there is only one height inside the area at a given position. This makes it possible to define the surface within the area as a set of three-dimensional points, where the first two-dimensional gives the position of the point and the third is the height of the surface at that position. These points are therefore connected with triangles to produce the surface using an intelligent triangulation algorithm. The retaining wall connects the separated regions together to form a finished three-dimensional surface structure.

The hierarchical solution process optimizes the surface shape such that hierarchical constraints are met and costs are minimized. Most hierarchical constraints are defined over a two-dimensional area either given by the planning solution process or typed in by the user. For example, parking lots are generally not allowed to tilt more than 4 degrees in any direction to prevent rolling of cars, but should be tilted at least 2 degrees to allow water to flow away. In addition to the minimum and maximum slopes, there are other constraints that affect the grade, with the most important perhaps being the minimum and maximum heights to prevent a grade for a certain area from falling below or above a specified height. The following provides a complete list of current constraints and how to enforce them. It is important to note that the optimization process as outlined below is independent of the individual constraints that make it possible to easily add new constraints.

(i) Huge search space

With all two-dimensional areas and the existing hierarchy provided by the user, this structure remains substantially unchanged after the surface is constructed. The hierarchical resolution process only changes the height of all three-dimensional points. By moving the point up and down, all triangles connecting the point will get a different slope. If the point moves downward, the triangle will tilt more toward the point, and if it moves upward, the triangle will tilt a little further away from the point. Also, there are many different surfaces with different slopes and heights with different possible costs. Because no "minimum variation" is defined, there are an infinite number of possible heights for each point in the surface.

This is not usually a problem for evolutionary algorithms. The EA can solve the problem of having an infinite search space with certain minimum and maximum values defined. Points of the existing surface are defined for the hierarchical resolution process either from contour lines or survey points from a user-entered system. To accurately input an average size surface, about 5000 points may be required. Some finer surfaces may require up to 10000 dots. Each of those points may vary independently. To illustrate this enormous search space, consider what happens if every point in the surface may have 10 different heights. This would give (on average) 105000 different surfaces (1 with 5000 zeros). For comparison, the number of atoms in the universe was estimated at 1081 (1 with 81 zeros).

The behavior of the evolutionary algorithm in a search space of this size is rarely known, and while it should be possible in theory, the present system does not provide the time required for the algorithm to learn the directional properties of this search space. Instead, this hierarchical resolution process should preferably send out answers in less than 5 minutes.

(ii) Local optimization

Even if the search space is large, two main features that make feasible optimization ranking are:

changing the points on the grading surface affects only the surrounding triangles and the connecting retaining walls; it does not change the properties of the other end of the site.

The existing ranking (although perhaps not feasible) is the lowest cost ranking to be generated. No earthworks need to be moved and no retaining walls need to be constructed. The cost of the earthwork is largely linear with the distance between the proposed grading and the existing grading.

These two points lead to the following assumptions:

if the triangle is tilted too much towards a certain point, moving that point upwards makes the site more feasible.

If a triangle tilts too far away from a point, moving that point down makes venue movement feasible.

Moving the point towards the original height makes the site less costly if it reduces the difference between the total segmentation and the total filling on the site.

Moving a point in the retaining wall towards its opposite part, reducing the size of the wall, making the site less costly.

From the above point of view, global optimizations that are difficult to solve are now defined from local optimizations that are easy to solve. Taking these rules and applying them in a deterministic, locally optimal form by adding another hypothesis is a very small step: all points may be changed simultaneously without changing any of the assumptions listed above. This effect is negligible when this assumption may not be exactly correct (since a change in one point does affect the triangle around it, and therefore the point affects the points on the other ends of the triangle) but a given change is not important. This type of optimization is used in Cellular Automata (CA) and is called "synchronous update". This method is in contrast to CA in that there are neighborhoods and points are influenced by the high degree of variation in their neighbors in the previous generation, but the surface structure is different and global influences such as total segmentation and fill balance are also unique, as is the application to earthwork calculations in civil engineering.

FIG. 19 illustrates this simple optimization process (feasibility and cost calculations omitted for purposes of illustration)

(iii) Terminate

The hierarchical resolution process needs to be terminated when the surface is feasible and the cost is optimal. To achieve this result, the hierarchical resolution process assumes the following:

the cost is optimal if the cost of the whole site does not vary by more than 0.5% after the last 1000 iterations.

This result is as good as the feasibility obtained if the feasibility of the site has not changed by more than 0.5% after the last 1000 iterations. If this result is not feasible, the constraints of the site cannot be fulfilled.

Although the above does not necessarily hold for each location, in practice these settings are very suitable for most locations. These settings can be modified if the required quality is high, but this will affect the runtime of the resolution process.

C.Utility solution process

The utility solution process is responsible for all pipeline issues. There are three main categories of situations where underground piping is required:

(a) drainage pipeline for underground water

(b) Life sewage pipeline

(c) Drinking water pipeline

From these three points, optimizing a drainage plan is relatively complex and also depends largely on the classification of the proposed site. Also, drainage optimization is an important feasibility test in terms of measuring the quality of the proposed grading.

(i) Utility (drainage) solution process

Like the other two solutions, the utility solution process can be divided into a simulation part and an optimization part. The simulation portion may also be divided into a plurality of portions, including: water flow calculation and pipeline throughput calculation. Both of which will be explained in detail below.

The optimization part in the utility solving process is composed of the optimization planning and the size of a pipeline connected with a storm rainwater inlet, a pool and a joint point. Smaller or shorter lengths of piping mean less cost to the site, but piping planning must be able to handle certain storms on the proposed scale as the scale resolution process is designed. Again, the constraints on this optimization can be quite complex, but should be accurate in order to prevent flooding of the designed site.

The height of the pipe is also constrained. Drainage pipes generally use gravitational forces to move water. Thus, the pipe has a minimum slope for water to flow. The pipes are not allowed to be above ground with a maximum depth below which they are damaged or hit. Furthermore, the duct is not allowed to pass through a certain area like a building, but is only rarely allowed to pass out of the property boundaries. All of these (and other) constraints make this a heavy "bounded" optimization and make it difficult to find an effective "optimization path" for the evolutionary algorithm.

A typical site may have about 20 entries/tap points spread throughout the site. This means that the number of possible pipes between the two inlets is 20 × (20-1)/2 or 190. The number of possible combinations of these pipes to form different plans is 2190, which is approximately 1,6 · 1057(1 followed by 6 and 56 0's). Also, it is not practical to carefully examine each different pipeline plan. Optimization without any assumptions in this situation would be difficult.

There are two assumptions that make this problem somewhat easier:

the cost of the piping must not be negative.

The best pipe plan for a subset of entrances will be part of the best pipe plan for the entire site.

The first assumption is important for the second assumption to hold. Civil engineering often uses a second assumption, which theoretically exists that this is not the case. This, along with other constraints, such as "need to pipe somewhere", for example, translates this problem into local optimization. The system knows where the water needs to go (i.e., to the sink/joint) so as to limit the number of possibilities for the initial of the problem, or alternative solutions. Also, because the solution process can assume that any pipe starts out well and ends well, the optimization becomes adding the "best" feasible pipe to the already added pipe, and the process is repeated until all of the inlets are connected together.

FIG. 20 illustrates an optimization cycle for generating a drainage plan for a site. Some similarity to the evolutionary cycle is noted. One difference is that for each iteration of this loop, the remaining problems are reduced, since it is not necessary to consider each entry connected to the pipeline map again.

The location where the portal is generated can be complex. This involves simulating the water flow and then finding where the water on the sidewalk surface becomes a pond. The system then checks whether the inlet is able to effectively manage/drain and if not, the system adds the inlet to a strategic location where water is captured prior to the aquatics pond. This simulation is performed using a flowsheet.

The following describes how to "consider adding a pipe". The addition of the pipe affects not only the inlet and the pipe itself, but also the flow of the added water through the pipe to the basin/fitting. This means that each pipe is tested by temporarily adding the pipe to the pipe map and then recalculating the feasibility, size and cost of all pipes. Adding a pipe when nothing in the piping diagram becomes infeasible because a pipe becomes infeasible, and is the least costly pipe to add; the concept is to add the least costly pipe at a time, giving a pipe map with the least cost at the end. When all the inlets are connected together, the cycle terminates.

(ii) Sewage and water supply optimization

Each building typically has one sewage line and one potable water line. These lines are usually connected in a straight line to the joint in a pipe. Optimization looks for the lowest cost, feasible joint point for each pipe. The gravitational force only works on the domestic sewage pipes, since drinking water uses pressure pipes.

Because sanitary sewer pipes use gravitational forces, the height of the building may have an impact on the feasibility of the drain. Although this is considered in the hierarchical resolution process. The building is not allowed to be below a certain height calculated using the lines on the sanitary sewer joint building and their relative positions.

Details of the calculation

The following discussion outlines the simulation and calculation of the cost model in the optimization engine.

A.Cost model

In an exemplary embodiment of the invention, the cost model is the heart of the optimization process. Everything is optimized according to cost.

The cost model has two types of inputs:

unit cost, the cost of all parts of the site.

Quantity, size/weight/length of the part of the site

(i) Unit cost

The unit cost used is as follows:

and (3) grading solution process:

(ii) measuring

For all these costs a measure is listed which needs to calculate the total cost of the site. The following discussion will outline how segmentation and padding are computed.

List of all measurements used in the cost model:

and (3) grading solution process:

and (3) utility solution process:

the cost model is mainly recalculated during the grading and drainage resolution process. The planning solution process will focus more on the number of parking spaces and the size of the paved area. However, the paving area does have a direct relationship with the paving area cost and thus with the planning cost. However, because this is only one aspect of the complex cost report, the entire cost report need not be recalculated in the planning solution process.

The entire cost model is recalculated in the following manner:

v/calculate all hierarchical costs

calculateClearing(VdisturbedArea，Vtopsoil)；

calculateCutAndFill(VearthCut，VrockCut，VunsuitableCut，Vfill)；

calculateWalls(VretainingWall)；

calculateFinish(VdisturbedArea，Vpaving，Vcurb，Vsidewalk)；

calculateErosionControl(VdisturbedArea)；

V/calculate all utility costs

calculateSWSInlets(VinletsEA，VinletsOversizedEA，

VinletsFT，VinletsOversizedFT)；

calculateSWSManholes(VmanholesEA，VmanholesOversizedEA，

VmanholesFT，VmanholesOversizedFT)；

calculateSWSRest(Vriprap，Vpumps，Vpondkits，Vstormmains)；

Note that the only cost to compute the planning solution process during the "calculated finish" function.

The different grading costs are calculated in the following way:

calculateClearing(VdisturbedArea，Vtopsoil)：

cost+＝VdisturbedArea＊Cclearing；

cost+＝VdisturbedArea＊Vtopsoil＊Cstripping；

calculateCutAndFill(VearthCut，VrockCut，VunsuitableCut，Vfill)：

Vonsite＝min(VearthCut，Vfill)；

Vwaste＝(VearthCut-Vfill)＊Fexpansion；

Vborrow＝(Vfill-VearthCut)/Fcompaction；

if(Vwaste＜0)Vwaste＝0；

if(Vborrow＜0)Vborrow＝0；

cost+＝Vonsite＊Cearth+VrockCut＊Crock+

Vunsuitable＊Cunsuitable；

cost+＝Vonsite＊Cfill；

cost+＝Vborrow＊Cborrow；

cost+＝Vwaste＊Cwaste；

calculateWalls(VretainingWall)：

cost+＝VretainingWall＊CretainingWall；

calculateFinish(VdisturbedArea，VpaVing，Vcurb，Vsidewalk)：

cost+＝Vfinish＊CdisturbedArea；

cost+＝Vpaving＊Cpaving；

cost+＝Vcurb＊Ccurb；

cost+＝VsideWalk＊Csidewalk；

calculateErosionControl(VdisturbedArea)：

cost+＝Verosion＊Cerosion；

the same procedure is performed for all utility correlation functions:

calculateSWSInlets(VinletsEA，VinletsOversizedEA，

VinletsFT，VinletsOversizedFT)：

cost+＝VinletsEA＊CinletsEA；

cost+＝VinletsOversizedEA＊CinletsOversizedEA；

cost+＝VinletsFT＊CinletsFT；

cost+＝VinletsOversizedFT＊CinletsOversizedFT；

calculateSWSManholes(VmanholesEA，VmanholesOversizedEA，

VmanholesFT，VmanholesOversizedFT)：

cost+＝VmanholesEA＊CmanholesEA；

cost+＝VmanholesOversizedEA＊CmanholesOversizedEA；

cost+＝VmanholesFT＊CmanholesFT；

cost+＝VmanholesOversizedFT＊CmanholesOversizedFT；

calculateSWSFES(Vfes)：

for every different size of FES do：

CthisFES＝cost in Cfes with current size；

VthisFES＝number of FES’s used with this size；

cost+＝CthisFES＊VthisFES；

calculateSWSRCP(Vrcp)：

for every different size of RCP do：

CthisRCP＝cost in Crcp with current size；

VthisRCP＝total size RCP pipe used with this size；

cost+＝CthisRCP＊VthisRCP；

calculateSWSRest(Vriprap，Vpumps，Vpondkits，Vstormmains)：

cost+＝Vriprap＊Criprap；

cost+＝Vpumps＊Cpumps；

cost+＝Vpondkits＊Cpondkits；

cost+＝Vstormmains＊Cstormmains；

the total cost of the site described above is calculated by the optimization engine. To optimize this cost, the solution process incrementally reduces this cost using evolutionary methods by changing different input parameters.

B.Planning a solution process

The planning solution process generates different feasible site plans and finds the "best". The optimal plan definition depends on what other resolution processes are stacked on this resolution process and certain specified constraints on the planned resolution process:

no stack of processes and no parking space with minimum number: the solving process finds the maximum value of the number of parking spaces that gives the boundary constraint.

No solution to process stacking and minimum number of parking spaces: the solution seeks to minimize the amount of parking space required.

The hierarchical resolution process is stacked on the planning resolution process with an optional utility resolution process: the solution process looks for the least costly plan in terms of total project cost.

The planning solution process generates a plan for each component location it may plan. In one embodiment, the planning solution locates only one building and then builds the site around that building. In other embodiments, the planning solution may locate multiple buildings and areas such as pools, waste bins, etc. For the planning solution process, these areas are indistinguishable and their use is defined by constraints present in the areas.

In addition to the active area, the planning solution processes also handle the fixed area. These space constraints, such as for example, ancillary buildings, may be taken into account.

A vehicle lane is defined from the center point. The planning solution then generates a motorway to the parking lot or building, depending on the central point and the constraints defined on the building.

(i) Locating a building

Buildings can be located in two ways:

using evolutionary cycles

-using iterative loops

(ii) Defining sidewalk

The walkway may be added to the building by the user. It becomes attached to the building and moves and rotates relative to the building. Each side of the building has a predefined walkway width. This default value is 0.

(iii) Defining a parking lot

The term "parking lot" relates to an area where parking is allowed, which in the case of planning a resolution process is translated into: area where parking is maximized. Thus, if a minimum amount of parking space is not set for a building, the parking lot is the boundary of the entire ownership minus the parking back and the ancillary buildings within the boundary. Fig. 21 provides an example of such a parking lot.

It is noted that in some cases the parking lot is divided into a plurality of blocks by a parking back or an affiliated building. This is considered feasible, but may result in unreachable parking spaces.

(iv) Planning out motor vehicle lanes

A motorway is defined as a plurality of lines (and including, for example, the number of entrances/exits, curb radii, corner radii, sidewalk types, and stacking distances) in which vehicles retreat. The vehicle lane always has a "source" and a "target". The source may be either the center point of a vehicle lane or another vehicle lane and the target may be either a parking lot or a side area with a direction. This direction is either "parallel" or "perpendicular" when describing the edge of interest to which the path connects.

The source can be connected to the target in several different ways, but the planning solution process attempts to find the least costly alternative.

If the vehicle is located in a parking lot, the area of the parking lot needs to be divided. Note that this may again be advantageous for all very feasible multi-block parking lots.

(iv) Planned parking partition

The resulting parking lot is "filled" with parking space. The idea is to maximize the amount of parking space. This is done by maximizing the amount of space in each parking lot individually. Some heuristics/assumptions are used to get this result:

-a boundary maximizing space of the parking lot of the parking sign.

Parking maximizes the space compared to buildings.

Parking sign boundaries should be in front of parking around the building

The following algorithm achieves these results:

for each parking lot:

generating parking bays from parking lot boundaries

Generating parking bays from the boundaries of the secondary building

Generating aisles to reach all parking bays with cross-distance

Creating interior parking bays

To generate an interior parking compartment, the planning solution process must have a direction and a starting point for the compartment. Alternatively, given the direction of the compartment, but if no one of the sides of the hypothetical parking lot is given, then the following algorithm is given to form the compartment:

for each side of the parking lot:

when there is a parking lot on the left:

separating parking partition from parking lot by using 2P depth

The aisle is divided from the parking lot by the width.

Calculating parking spaces

Partitioning based on parking space using optimal edge planning

(vi) Planning a parking space

To plan the parking space, the solution takes into account the following constraints:

the width and depth of the parking space need to be sufficient.

Roadside need to consider safety islands.

-parking spaces that need to be accessible.

Parking spaces may have to be aligned.

The parking space is created by inserting the safety island and offset portions of the ends to create useful portions of the bay, using the boundaries of the parking bay. These useful portions are then considered according to the size and then "forming" the rows of the parking space.

C.Hierarchical resolution process

The hierarchical resolution process seeks to optimize the proposed site surface in a manner that minimizes the overall site cost. This is done using a local heuristic evolutionary approach to change the proposed surface.

The surface of the system is defined by triangles. These triangles are generated using a combination of "Delauney triangulation" and boundary "shaping". The following statements are important to the hint generated data structure:

-each point has a list of connected triangles

Each triangle has three distinct points

-each point has a current and original altitude

Some points are "linked" to other points above or below to form a wall

The hierarchical resolution process may be run alone or in combination with the utility/drain resolution process. If the staging solution process is run in conjunction with the utility solution process, it seeks to minimize the inlets and lift and lower the pool and inaccessible inlets to make drainage feasible and cost effective.

How the hierarchy resolution process optimizes the hierarchy, details and costs regarding the determination of feasibility are discussed further below.

(i) Feasibility

There are many feasibility constraints that need to be implemented in order to obtain a viable hierarchical solution. These constraints must be translated into properties for the point. There are two different types of hierarchical constraints:

-ramp constraints

-height constraints

The ramp constraint constrains the slopes of the triangle between the minimum and maximum ramps. This is typically defined for the entire region, but may also be defined for a virtual region defined in terms of an offset of an edge or point. Important ramp considerations include:

slope of the entire area

Slopes far from the building

Slope inside a building in a certain direction

Slope at the curb

To implement the ramp constraint, the triangle that violates the constraint needs to be tilted or leveled. This can be translated into moving points away from the average height if tilted and changing points closer to the average height if leveled.

Height constraints are also typically defined over the entire area, but may be dynamic in nature. Thus, the minimum or maximum slope may depend on the height of another region or point. This defines for example the different parts of the pool possible.

Due to the dynamics of certain high constraints, points that violate minimum or maximum constraints do not move immediately to a particular constraint, but rather slowly toward the constraint.

One particular height constraint is the maximum retaining wall height. This is in fact the maximum height difference between the two connection points.

(ii) Cost of

One of the main objectives of the hierarchical resolution process is cost optimization. However, the resulting solutions are only relative if they are also feasible. Also, it will do so if it improves feasibility, regardless of the cost of certain changes. Each point tracks feasibility improvements and cost improvements. Only when the feasibility improvement is close enough to 0, the cost improvement is also applied.

Changing the height of a single point affects cost in several ways, including:

-cutting/filling earthwork and rock

Output/input earthwork

-disturbing the earth

Height of retaining wall

Number of inlets

Some of the above mentioned items can be computed locally and others have to be processed in a global manner for each iteration. The cost calculation per iteration is generally performed as follows:

computing total input and output

Calculating the number of entries

...

For each point, the following is performed:

...

if the point is feasible, then:

calculating the impact of changes on mill and rock segmentation

Calculating the impact of changes on a perturbed plant

Calculating impact on optional retaining walls

Calculating directions of balanced input/output

Calculating whether an entry should be removed

...

For each point, the following is performed:

change point

(iii) Feedback cycle from drainage

One possible drawback of using local search algorithms and heuristics that rely heavily on local orientation is the orientation required to optimize each aspect. The problem is that drainage is difficult to define in terms of local properties. The number and size of the inlets may give some feedback, but for example the height of the pool is not as easily defined.

Currently, the height of the pool is defined as the average of the defined area. After approximately 500 iterations of the hierarchical resolution process, 1 drainage iteration is performed. The results of the utility/drain resolution process are then considered to be optimal for the given pool height. The height of the pool (variance) is then changed and the hierarchy is run for another 500 iterations, with an additional single run of the utility solution process. If the result is improved, the new height is used for the next mutation, or the old one is used. This process is very similar to the (1 + 1) Evolutionary Strategy (ES)

D.Utility solution process

The utility resolution process can be divided into three parts:

(a) storm sewer

(b) Domestic sewage sewer

(c) Drinking water

The implementation of the present system will address the following utility problems:

-flow of water

-drainage area location

-inlet position

Optimization of the pipeline

Certain sanitary sewer and drinking water constraints

To find the lowest points in the classification and how much water to drain into these lowest points, the utility solution process must simulate the flow of water. In one embodiment, a cellular automaton is used to simulate the flow of water in the form of a grid. The second algorithm uses a "flow graph" to simulate a more perfect directional route. The first algorithm can be used as a visual tool, and as a tool for flood prediction. A second algorithm may be used to calculate the location of the inlet and the size of the drain area.

(i) Cellular automaton

Two-dimensional cellular automata were used to simulate the flow of water. Such cellular automata works by defining a small (in this case) rectangular block of the whole area of the cell, which has a certain fraction of the solution in the form of values. In this case, each cell has the total amount of water currently present in the cell. The idea is therefore to define rules for how the solution (in this case water) interacts with neighboring cells. In a synchronous cellular automaton, this is done using a set of rules by recalculating values in the cells for each iteration using values of neighboring cells in the previous iteration. These rules then define the behavior of the CA, along with the topology of the CA.

The neighborhood of cells may be defined in different ways, like for example the "von Neumann" (von Neumann) neighborhood. In this neighborhood, two cells are neighbors if and only if they are directly connected horizontally or vertically. That means that there are four neighbors per cell.

The simulation of the water flow by the algorithm in this CA is based on a triangular slope of cell overlap. Each cell is accurately given a triangle in the center of the cell. The slope of that triangle will therefore represent the flow of all the water in the cell. This makes an estimate of this algorithm and it may be inaccurate because the cells are too small. Cells can be adjusted to any size and the level of accuracy is limited to exceeding the level of data accuracy itself in the surface.

The standard update cycle is similar to:

for all cells, the following was performed:

water in cells =1.0

While there are still changes:

for all cells, the following was performed:

flow to cell =0

For all cells, the method is carried out

Slope in x = x direction

Slope in y = y direction

len = length of the ramp vector

If x >0 then:

flow to the right neighbor + = water x/len

Otherwise, carrying out:

flow to left neighbor + = water x/len

If y >0 then:

flow to upper neighbor + = water x/len

Otherwise, carrying out:

flow to the lower neighbor + = water x/len

For all cells:

water in cells = flow to cells

(ii) Flow chart

Flow graph algorithms attempt to capture the total flow of surfaces in a graph structure. Once that map is generated, it is easy to calculate basic functions like the flow direction to the inlet, the delineation of the drainage area and the size of the drainage area.

The flow graph is generated using the "nodes". Each triangle in the surface gets one of these nodes. Each triangle in the data structure has knowledge of its neighbors and even its neighbors (or "connected triangles") in its neighborhood. A node may be connected to other nodes if the triangular water drains into an adjacent triangle. Each such connection is called a "flow" and all connections have a certain "percentage" of the total flow of the triangle assigned to it. This total flow percentage is calculated as the percentage of triangles that flow from the edge to the adjacent triangle according to the slope.

The drainage is calculated at the end of the grading, thus requiring the roadside to be considered. The curb stops water flow in one direction but allows flow in the other direction. Also, water can flow from the sidewalk into, for example, a parking lot, but not into other paths. The data structure defines the region boundaries in such a way that each triangle is only on one side of the boundary, so no boundary crosses a triangle. Likewise, the neighbors of a triangle on the other side of the boundary are connected to neighboring triangles on the same side of the boundary in a different manner.

The type of area defines whether the curb is used to define the area boundary. Typically, roadsides are used only where paved areas meet non-paved areas. Each region has a "needscutting" flag that determines this usage. For a flowsheet, this means that no flow is allowed from paved areas to non-paved areas. Allowing flow in the other direction. Any boundary that does not allow water to pass through is treated as if it were a channel. That means that water will flow to the lowest end of the edge that will divide the water block and that a helper node will be created.

Retaining walls are also water flowing stops. Water is not allowed to flow from the lower region to the upper region. This is handled in a roadside-like manner. However, if flow through the retaining wall is allowed (from higher to lower), a flow of water from a high point to a low point is generated. This precedes any other flow from the high point to the channel triangle.

(iii) Drainage area

A drainage area is the general area of the surface that drains water to some local minimum. The local minimum can be easily found completely by searching all nodes for nodes without any outflow. Theoretically each local lowest node has a drainage area.

To calculate the size of the drainage area, the algorithm starts at the local lowest point and recursively goes to "up-flow" and adds the relative scale of the flow to the same amount of water as the size of the drainage area. This is similar to the following:

the function of calculating the flow size into the node:

if the size has already been calculated

Size of return

size=0

If the node has a triangle, then:

size + = size of triangle

For all nodes of the ingress node:

nsize = calculating the flow size into the source node

size + = percentage of flow. size = size $

Size of return

Because each triangle has an area and each area has a runoff (run off) coefficient, the runoff coefficient for a drainage area may be calculated in a similar recursive manner.

The drainage area is delineated by the scoring between triangles where most of the flow goes to different inlets. When this is an estimate over the estimate, it is mainly used for visualization purposes and is therefore sufficiently accurate.

(iv) Inlet port

The standard inlet is placed at the local lowest point of the drainage area. Not all inlets require pipelaying. All drainage inlets would otherwise be located in unaffected areas where pipelaying is required. The unaffected areas are all paved areas plus all constructed areas. All impermeable areas outside the local nadir do not need to be drained and should actually be drained in a manner similar to that before the site grading.

The land for sale (out parcel) is an exception to this rule. The land for sale is classified into the intended buildings and parking lots, but they are not paved yet, although most of the land for sale requires drainage.

Some drainage areas are too large for only one inlet. If the drain area exceeds the maximum size of the inlet, more inlets need to be added to the drain area. Additional entries are added by proceeding up the flow graph from the existing entry until the flow to the current location is less than the maximum entry size. That process is similar to:

when the inlet is too large, it is carried out:

Current=inlet

when the current flow is too large:

go up to the node with the largest flow

Adding a current entry

Recalculating flow to an entry

This algorithm does assume that no triangle is larger than the maximum size of the entry, but can ensure this by dividing the triangle during triangulation.

(v) Pipeline

The process of generating the pipeline uses a greedy algorithm. To generate the pipeline map, the algorithm uses all of the entries generated at earlier times plus all of the junction points. The connection point may be either a point on an existing pipeline or a sink. The algorithm then generates all possible conduits between the inlet itself and the junction point. Piped from the inlet to the end of the pool in the closest point of the pool surface boundary. Some of the conduits are of course out of specification. For example, plumbing through a building is not regulated. Those pipes are arranged around the building in the shortest possible way. This is also a result of pipeline efforts to cross ownership boundaries.

For each combination of inlets, the following is performed:

generating a pipeline

For all obstacles, the following were performed:

passing the pipe around the obstacle

For all joint points, proceed

For all entries, the following was performed:

generating a pipeline

For all obstacles, the following were performed:

passing the pipe around the obstacle

After this pre-treatment, there is a list of all the eligible pipe paths from each inlet to each other inlet and from each inlet to each junction point.

To generate a pipeline graph from a list of all possible pipelines, a greedy algorithm adds the "best" pipeline to an already existing graph. The optimal pipeline is defined here by the minimum cost of the entire pipeline graph. That means that the cost of the entire pipeline graph needs to be calculated each time a pipeline is "tried". Water needs to flow through the pipe by gravity. That means that water flowing to the inlet at a certain height cannot flow out of the inlet at a higher height. This plus the requirement that each duct be between certain minimum and maximum slopes means that if one duct is added, certain heights may vary.

The pipe also needs to be "resized". That means that the amount of water flowing through the pipe per second will indicate the minimum size required to make the flow feasible. This is done using the Manning equation. This equation estimates flow through the pipe using the size of the area drained into it, the time for the water to reach the pipe and things like a slope, and the material used in the pipe itself. There is a standard set of pipe sizing and materials used, and the algorithm selects the most cost effective one for the current pipe map. The algorithm for generating the pipeline map is similar to the following:

graph = set of linker points

When all entries are not completed:

for all the pipes next to the current graph, we proceed to:

connecting a pipe to a figure

Recalculating cost of a graph

Disconnecting a pipeline

Connecting the cheapest conduit to the figure

Evolutionary algorithm

Evolutionary Algorithms (EAs) evolve a set of different, either globally optimal or locally optimal solutions-each solution conceptually satisfying cost measures (on-site) in a cost-effective manner, taking into account system and user constraints and user preferences.

Starting at the "0" generation, the first step in the EA is to create an initial population of conceptual solutions. Each solution includes a different set of parameters that drive the optimization. This initial population may contain any of a single to thousands or more of the potential solutions.

For each solution in the population, a cost measure is defined. For example, grading costs include aspects such as, for example, total distribution area, total volume of excavated material, volume of excavated rock, volume of excavated unsuitable material, volume of infill material, retaining wall area, parking area, concrete walkway area, length of curbs and gutters, and sloped surface area.

Referring to FIG. 22, after the initial population is created, the next step is to apply a fitness function to quantitatively estimate the fitness of each candidate solution. This step involves first determining the engineering feasibility of the solution and whether the solution satisfies the selection rules discussed above. If the solution meets these threshold requirements, then a score for fitness is assigned using the cost model and any applicable penalties. If not, the solution is discarded immediately. Instead of discarding such a solution, the method may also provide measures to avoid creation of such a solution altogether, or to facilitate repair of such a solution, both measures being based on avoiding or repairing with heuristics.

For those solutions that meet the above threshold requirement, fitness values are assigned to the cost measures. In this example, this fitness value is a measure of the cost of the current solution. As stated previously, cost penalties are assigned to measures that violate user preferences or "soft constraints". The cost for each measurement is calculated based on a cost model.

For each solution in the population, a cost and penalty are added to derive fitness values in a manner that introduces a weight factor for the cost as well as various penalty components. This can be expressed by the following equation:

fitness＝w_c*cost+w_p1*penalty₁+w_p2*penalty₂+…+w_pn*penalty_n

where "cost" refers to the value of the cost function and "dependency_i"refers to the value of the penalty cost value for the infeasible component of the current solution. This combines the cost and penalty factors into a single fitness value, which needs to be minimized. In addition, the cost can be kept aloneAnd penalty values and for estimating the quality of the solution in different ways, e.g. by defining a specific order over (cost, penalty) pairs or by considering cost and penalty as separate criteria in a multi-objective optimization task, characterized in that the conflicting optimization criteria are used to determine the so-called Pareto-front of the best possible compromise solution between the conflicting criteria.

After scoring each solution in the population, the EA determines whether known termination criteria are met. In this example, the termination criteria is a preselected number of rounds or "generations". Assuming this criterion is not met, the system selects certain candidate solutions to replicate into the population of offspring. The EA can do this using many different techniques; i.e. choice of elite, choice of fitness scale, roulette choice, scale choice, competition choice, ranking choice, fertility choice, steady state choice, tier choice, (mu, lambda) -and (mu + lambda) -choices (also called cutoffs). Some of these methods are mutually exclusive, but others may be and often are used in combination. There may also be two selection steps, one often referred to as "sex selection" (i.e., making individual copies from parent populations) for breeding, and another referred to as "environmental selection" for reducing the size of offspring populations.

According to elite selection, it is guaranteed that the most adapted solution per generation will be selected. In fitness ratio selection, more suitable individuals are more likely, but not necessarily, to be selected. Roulette selection is a form of fitness proportion selection in which a solution is selected in an opportunity proportional to the amount by which its fitness is greater or less than that of its competitor. According to scaling selection, as the average fitness of the population increases, the intensity of the selection pressure also increases and the fitness function makes the distinction clearer. This approach may be helpful in making the best choice later when all solutions have relatively high fitness and have little difference in fitness from each other. In competition selection, a small number of subgroups in a solution are selected randomly and repeatedly from a larger population and the members of each subgroup compete with each other. Only one of the propagated from each subgroup-i.e. the best, solution-was selected. In the queuing selection, each solution in the population is assigned a numerical order based on fitness, and the selection is based on this order rather than the absolute difference in fitness. The advantage of this approach is that it can prevent well-fitting individuals from gaining advantage earlier at the expense of less-fitting individuals, which can reduce population genetic divergence and can hinder attempts to find an acceptable solution. In fertility selection, the offspring of the solution selected from each generation becomes the whole next generation. No solution is reserved between generations. In steady state selection, offspring from the solution selected in each generation are traced back to the preexisting population, replacing some of the less suitable members of the previous generation. Some solutions remain between generations. In hierarchical selection, the solution scrutinizes multiple rounds per generation of selection. Lower grades evolve faster and less discriminately, while those existing at higher grades are more relentless. The advantage of this approach is that it reduces the overall computation time by eradicating most solutions that show little or no desire with faster, less selective evolution, and subjecting only those surviving from this initial test to more rigorous and computationally expensive fitness evolution. In (μ, λ) -selection, the best solution is definitively selected from the λ solutions in the offspring population (in this case λ is greater than μ) to form the parent population for the next iteration of the evolutionary algorithm. The advantage of this approach is that it is simple and supports the adaptive capability of the evolutionary strategy, and it allows the algorithm to move from local optimal towards globally optimal solutions. In (μ + λ) -selection, μ best solutions are deterministically selected from μ solutions in the parent population plus λ solutions in the offspring population to form the parent population for the next iteration of the algorithm. The advantage of this method is that it is elite, i.e. it ensures that the solution does not deteriorate during the optimization process. The (μ, λ) -and (μ + λ) -selections are also order-based selection methods, both used in evolutionary algorithms as "environmental selection" methods (i.e., after the generation and evolution of new solutions-progeny individuals-have occurred).

In one example, an order selection method, such as (μ, λ) -selection, selects all μ candidate solutions that have the best (i.e., the smallest) cost and penalty function value among all λ solutions in the current offspring population.

Once the selection has chosen the appropriate solutions, they are randomly changed when it is desired to improve the fitness of their next generation. This random change occurs through mutation and crossover. The solution is to mutate by slightly changing individual parameters. Such a variation may also involve adaptation of one or more of variance and covariance of the distribution of the variation, such as, for example, a normal distribution. A possible example of adaptive variation is to change the coordinate vector x (x) with associated variance (standard deviation) by the following mathematical procedure_l，…,x_n)：

Generating a new value of σ, denoted σ' as σ ═ σ exp (τ × N (0,1)), cycling through all values i of i ═ 1, …, N and generating x_iNew value of (1), from x'_iIs represented by and x_i＝x_i+σ＇*N_i(0,1)。

Here, N (0,1) is a random number according to a normal distribution having a mean value of 0 and an expected 1.τ is a parameter of the method, which may be set to a value of 1/sqrt (n). Crossover requires the selection of two or more solutions to exchange one or more parameters, thereby producing artificial "offspring" that are a combination of their parents. Because of the crossover, there is successful information transfer between "individuals" -a solution that can benefit from what other solutions have learned, and plans can be mixed and combined, with the potential to produce offspring that have the advantages of its parents without their weaknesses. A common form of crossover, also known as discrete recombination in evolutionary strategies, called uniform crossover, allows each parent to have a 50% chance of contributing a parameter to the newly formed solution. An interleaved illustrative example is provided in fig. 23. In this example, parameters are exchanged between the two solutions (represented by "parent 1" and "parent 2") by a random exchange of values for each location that has a 50% chance to exchange or dwell to arrive at two new solutions (represented by "descendant 1" and "descendant 2"). Also, a continuous split of information can be exchanged after a breakpoint or between two breakpoints, as shown in FIG. 24 (where a random breakpoint occurs after position three). All of these crossover operators generate two or more new solutions, which may use either all of them or only one of them, and then generally randomly choose between the generated new solutions. As with the evolution strategy used in this example, one of the new solutions is selected. Also, averaging one or more parameters between two or more solutions is a possible form of interleaving used by the evolutionary algorithm. An example of crossover of this version (also referred to as intermediate recombination in the evolutionary strategy) is shown in FIG. 25. This operator usually generates only one new solution. Also, any other arithmetic combination of one or more parameters between two or more solutions may be used as a possible form of interleaving, in addition to averaging.

This evolutionary policy algorithm may also be combined with local heuristics. The local heuristic method ensures that operators of the evolution strategy, specifically mutation operators, consider local constraint conditions, so that feasible points are generated by the mutation operators.

Exemplary embodiments of the present invention are described above. No element, act, or instruction used in the description should be construed as critical or essential to the invention unless explicitly described as such. Various details of the invention may be changed without departing from its scope. Furthermore, the foregoing description of the exemplary embodiments and best mode for practicing the invention are provided for the purpose of illustration only and not for the purpose of limitation-the invention being defined by the claims and their equivalents.

Claims (30)

1. A computer-implemented land planning system designed to generate at least one conceptual cost-optimized solution to a user-defined land development problem, the system comprising:

means for electronically creating at least one candidate solution to the land development problem when executed by a computer, the candidate solution comprising a plurality of interrelated engineering cost measures applicable to development of an undeveloped land site, and the plurality of engineering cost measures selected from the group consisting of site layout, site classification, and site utility;

means for employing an iterative heuristic problem-solving strategy to process the engineering cost measures of the candidate solution until at least one cost-optimized solution to the land development problem is implemented, wherein a change in one of the plurality of engineering cost measures of the candidate solution affects a change in another of the plurality of engineering cost measures of the candidate solution; and

means for transmitting output data illustrating said cost-optimized solution to the land development problem to a user when executed by a computer.

2. A computer-implemented land planning system according to claim 1, and comprising means for accessing user preferences for undeveloped land sites when executed by the computer.

3. A computer-implemented land planning system according to claim 1, wherein said output data contains documents comprising at least one graph.

4. A computer-implemented land planning system according to claim 1, wherein said output data contains documents comprising a project-ized cost list of said engineering cost measures.

5. A computer-implemented land planning system according to claim 1, wherein said output data contains documents that are transmitted to users over a global communications network.

6. A computer-implemented land planning system according to claim 1 and comprising means for accessing land development constraints of undeveloped land sites when executed by a computer.

7. A computer-implemented land planning system designed to generate at least one conceptual cost-optimized solution to a user-defined land development problem, the system comprising:

means for electronically creating, when executed by a computer, at least one candidate solution to a land development problem, the candidate solution comprising a plurality of interrelated engineering cost measurements applicable to development of an undeveloped land site, and the plurality of engineering cost measurements being selected from the group consisting of demolition, cleaning, excavation, filling, wall hugging, erosion control, completion leveling, earth output, earth input, drainage, pond height, parking lot design, approach, rock pavement, roadside, drainage, asphalt, terrain, vertical culverts, access ports, pipes, pipe dimensions, pipe paths, storm water collection systems, public lavatory sewer collection systems, and potable water systems;

means for employing an iterative heuristic problem-solving strategy to process the engineering cost measures of the candidate solution until at least one cost-optimized solution to the land development problem is implemented, wherein a change in one of the plurality of engineering cost measures of the candidate solution affects a change in another of the plurality of engineering cost measures of the candidate solution; and

means for transmitting output data illustrating said cost-optimized solution to the land development problem to a user when executed by a computer.

8. A computer-implemented land planning system according to claim 7 and comprising means for accessing user preferences for undeveloped land sites when executed by the computer.

9. A computer-implemented land planning system according to claim 7, wherein said output data contains documents comprising at least one graph.

10. A computer-implemented land planning system according to claim 7, wherein said output data contains documents comprising a project-ized cost list of said engineering cost measures.

11. A computer-implemented land planning system according to claim 7, wherein said output data contains documents that are transmitted to users over a global communications network.

12. A computer-implemented land planning system according to claim 7 and comprising means for accessing land development constraints of undeveloped land sites when executed by a computer.

13. A computer program product containing program instructions tangibly stored on a non-transitory computer-readable medium and operable to cause a computer apparatus to perform a method designed to generate at least one conceptual cost-optimized solution to a user-defined land development problem, the method comprising:

electronically creating, using a computer, at least one candidate solution to a land development problem, the candidate solution comprising an interrelated plurality of engineering cost measurements applicable to development of an undeveloped land site, and the plurality of engineering cost measurements being selected from the group consisting of demolition, cleaning, excavation, filling, wall hugging, erosion control, completion leveling, earth output, earth input, drainage, pond height, parking lot design, approach, block-faced road, curb side, drainage ditch, asphalt, terrain, vertical flow culvert, access opening, pipe size, pipe path, storm water collection system, public lavatory sewer collection system, and potable water system;

processing, using a computer, the engineering cost measurements for the candidate solution with an iterative heuristic problem solution strategy until at least one cost-optimized solution to the land development problem is implemented, wherein a change in one of the plurality of engineering cost measurements for the candidate solution affects a change in another of the plurality of engineering cost measurements for the candidate solution; and

looking at a solution to said cost optimization for land development problems.

14. The computer program product of claim 13, and the method comprises accessing user preferences for undeveloped land locations.

15. The computer program product of claim 13, wherein reviewing the cost-optimized solution comprises receiving a document including at least one computer-generated graph.

16. The computer program product of claim 13, wherein reviewing the cost-optimized solution comprises receiving a document including a project-ized cost list of the engineering cost measures.

17. The computer program product of claim 13, wherein reviewing the cost-optimized solution comprises receiving a document over a global communication network.

18. The computer program product of claim 13, the method comprising accessing land development constraints for an undeveloped land site.

19. A non-transitory computer-readable storage medium storing computer-executable instructions comprising one or more instructions executable by processing logic of a computing device, the computer-executable instructions, when executed by the processing logic, causing the processing logic to perform a method designed to generate at least one conceptual cost-optimized solution to a user-defined land development problem, the method comprising:

electronically creating, using a computer, at least one candidate solution to a land development problem, the candidate solution comprising a plurality of interrelated engineering cost measurements applicable to development of an undeveloped land site, and the plurality of engineering cost measurements being selected from the group consisting of demolition, cleaning, excavation, filling, wall hugging, erosion control, completion leveling, earth output, earth input, drainage, pond height, parking lot design, approach, block-faced road, curb side, drainage ditch, asphalt, terrain, vertical flow culvert, access opening, pipe size, pipe path, storm water collection system, public lavatory sewer collection system, and potable water system;

means for processing, using a computer, the engineering cost measurements for the candidate solution using an iterative heuristic problem solution strategy until at least one cost-optimized solution to the land development problem is implemented, wherein a change in one of the plurality of engineering cost measurements for the candidate solution affects a change in another of the plurality of engineering cost measurements for the candidate solution; and

looking at a solution to said cost optimization for land development problems.

20. The non-transitory computer readable storage medium of claim 19, wherein the method further comprises accessing user preferences for the undeveloped land site.

21. The non-transitory computer readable storage medium of claim 19, wherein reviewing the cost-optimized solution comprises receiving a document including at least one computer-generated graph.

22. The non-transitory computer readable storage medium of claim 19, wherein reviewing the cost-optimized solution comprises receiving a document including a project-ized cost list of the engineering cost measure.

23. The non-transitory computer readable storage medium of claim 19, wherein reviewing the cost-optimized solution comprises receiving a document over a global communication network.

24. The non-transitory computer readable storage medium of claim 19, wherein the method further comprises accessing land development constraints for an undeveloped land site.

25. A computer-implemented land planning method designed to generate at least one conceptual cost-optimized solution to a user-defined land development problem, the method comprising:

electronically creating, using a computer, at least one candidate solution to a land development problem, the candidate solution comprising a plurality of interrelated engineering cost measurements applicable to development of an undeveloped land site, and the plurality of engineering cost measurements being selected from the group consisting of demolition, cleaning, excavation, filling, wall hugging, erosion control, completion leveling, earth output, earth input, drainage, pond height, parking lot design, approach, block-faced road, curb side, drainage ditch, asphalt, terrain, vertical flow culvert, access opening, pipe size, pipe path, storm water collection system, public lavatory sewer collection system, and potable water system;

processing, using a computer, the engineering cost measurements for the candidate solution with an iterative heuristic problem solution strategy until at least one cost-optimized solution to the land development problem is implemented, wherein a change in one of the plurality of engineering cost measurements for the candidate solution affects a change in another of the plurality of engineering cost measurements for the candidate solution; and

looking at a solution to said cost optimization for land development problems.

26. A computer-implemented land planning method according to claim 25, wherein said method further comprises accessing user preferences for undeveloped land sites.

27. A computer-implemented land planning method according to claim 25, wherein viewing said cost-optimized solution comprises receiving a document comprising at least one computer-generated map.

28. A computer-implemented land planning method according to claim 25, wherein viewing said cost-optimized solution comprises receiving a document comprising a project-ized cost list of said engineering cost measures.

29. A computer-implemented land planning method according to claim 25, wherein reviewing said cost-optimized solution comprises receiving documentation over a global communications network.

30. A computer-implemented land planning method according to claim 25, wherein said method further comprises accessing land development constraints for undeveloped land sites.