CN119940826A - A task scheduling method for intelligent warehousing system - Google Patents
A task scheduling method for intelligent warehousing system Download PDFInfo
- Publication number
- CN119940826A CN119940826A CN202510019644.3A CN202510019644A CN119940826A CN 119940826 A CN119940826 A CN 119940826A CN 202510019644 A CN202510019644 A CN 202510019644A CN 119940826 A CN119940826 A CN 119940826A
- Authority
- CN
- China
- Prior art keywords
- task
- robot
- warehousing
- tasks
- storage
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Landscapes
- Feedback Control In General (AREA)
Abstract
本发明涉及智能仓储系统控制技术领域，具体为一种智能仓储系统的任务调度方法。所述方法包括以下步骤：首先，根据订单信息、仓储机器人状态和仓储环境信息的实时数据，生成仓储系统环境地图和总任务池；然后，基于所述仓储系统环境地图和总任务池，构建任务调度模型；接着，基于所述任务调度模型，设计任务调度算法，将总任务池中的任务分配给各个仓储机器人；然后，针对每个仓储机器人的任务池，设计任务次序优化算法，并生成每个仓储机器人的执行计划；接着，当总任务池新增或取消任务时，基于当前情况重新分配总任务池中的任务；最后，对于仓储机器人动态故障的情况，调度中心将故障仓储机器人未完成的任务回滚到调度中心，重新进行任务分配。
The present invention relates to the field of intelligent warehousing system control technology, and specifically to a task scheduling method for an intelligent warehousing system. The method comprises the following steps: first, based on the real-time data of order information, warehousing robot status and warehousing environment information, a warehousing system environment map and a total task pool are generated; then, based on the warehousing system environment map and the total task pool, a task scheduling model is constructed; then, based on the task scheduling model, a task scheduling algorithm is designed to assign tasks in the total task pool to each warehousing robot; then, for the task pool of each warehousing robot, a task sequence optimization algorithm is designed, and an execution plan for each warehousing robot is generated; then, when a task is added or cancelled in the total task pool, the tasks in the total task pool are reallocated based on the current situation; finally, in the case of a dynamic failure of a warehousing robot, the dispatch center rolls back the unfinished tasks of the failed warehousing robot to the dispatch center, and reallocates the tasks.
Description
Technical Field
The invention relates to the technical field of intelligent warehousing system control, in particular to a task scheduling method of an intelligent warehousing system.
Background
With the rapid development of electronic commerce and logistics industry, intelligent warehousing systems gradually become key technologies for improving warehousing management efficiency and reducing operation cost. Traditional warehousing systems mainly rely on cooperative work of manual operation and mechanical equipment, and although logistics and warehousing requirements are met to a certain extent, a plurality of defects still exist in the aspects of efficiency, accuracy and flexibility. Especially under the condition of large order processing amount and various types, the traditional warehousing system is easy to have the problems of low efficiency, high error rate, high labor cost and the like.
The intelligent warehousing system can realize high-efficiency, accurate and flexible management of warehousing operations by introducing automatic, informationized and intelligent technologies. The intelligent warehouse system not only can automatically finish tasks such as storage, picking and carrying of materials, but also can improve the utilization rate and the working efficiency of warehouse resources through data analysis and an optimization algorithm. Wherein, the storage robot plays a key role as an important component part of the intelligent storage system. The storage robot can automatically navigate, avoid barriers and carry goods in the warehouse, so that the manpower demand and misoperation are effectively reduced, and the overall operation efficiency is improved.
However, intelligent warehousing systems face a variety of complex dynamic events in actual operation, such as order changes, equipment failures, environmental changes, and the like. These dynamic events place higher demands on the stability and efficiency of the system. Particularly when a task is dynamically changed or a storage robot fails, the system needs to respond quickly and redistribute the task so as to ensure the continuity and the high efficiency of the storage operation. Therefore, research on an intelligent warehousing system method capable of processing dynamic events and optimizing task scheduling has important significance for improving the robustness and adaptability of the system.
In order to solve the technical problems, the invention provides a task scheduling method of an intelligent warehousing system.
Disclosure of Invention
The invention provides a task scheduling method of an intelligent warehousing system, which is used for solving the technical problems. The method comprises the steps of firstly, generating a warehouse system environment map and a total task pool according to order information, warehouse robot states and real-time data of warehouse environment information, then, constructing a task scheduling model based on the warehouse system environment map and the total task pool, wherein the task scheduling model comprises task set and robot set description, path constraint, resource constraint and task priority constraint, takes minimized path expenditure, task completion time and robot energy consumption as optimization targets, then, designing a task scheduling algorithm based on the task scheduling model, distributing tasks in the total task pool to each warehouse robot to obtain a task pool of each warehouse robot, then, designing a task order optimization algorithm based on the task pools of each warehouse robot, optimizing task execution sequence according to the distance between a task starting point and a task ending point and task importance, generating an execution plan of each warehouse robot, then, monitoring the warehouse environment and the task states in real time in the task execution process, namely, adding tasks and canceling tasks based on the current task execution conditions, re-distributing tasks in the total task pool, finally, completing the task scheduling center to the warehouse robot dynamic fault scheduling center, and the task to the warehouse system according to the current fault scheduling situation, and re-distributing tasks to the warehouse system.
A task scheduling method of an intelligent warehousing system comprises the following steps:
s1, generating a warehousing system environment map and a total task pool according to order information, a warehousing robot state and real-time data of warehousing environment information;
s2, constructing a task scheduling model based on the warehouse system environment map and a total task pool, wherein the task scheduling model comprises task set and robot set description, path constraint, resource constraint and task priority constraint, and takes minimized path overhead, task completion time and robot energy consumption as optimization targets;
s3, designing a task scheduling algorithm based on the task scheduling model, and distributing tasks in the total task pool to each storage robot to obtain a task pool of each storage robot;
s4, designing a task sequence optimization algorithm aiming at a task pool of each storage robot, optimizing a task execution sequence according to the distance between a task starting point and a task ending point and the importance of the task, and generating an execution plan of each storage robot;
s5, monitoring the storage environment and the task state in real time in the task execution process, and reallocating the tasks in the total task pool based on the current task execution condition when the tasks are newly added and cancelled in the total task pool;
S6, for the situation of dynamic faults of the storage robots, the dispatching center rolls back the tasks which are not completed by the faulty storage robots to the dispatching center, and then task distribution is carried out again according to the current storage system environment map and the updated total task pool.
Further, generating the warehousing system environment map and the total task pool according to the order information, the state of the warehousing robot and the real-time data of the warehousing environment information, and constructing the warehousing system environment map by adopting a grid method, wherein the specific steps comprise:
The method comprises the steps of constructing a two-dimensional rectangular coordinate system of a storage environment, wherein an x-axis represents the length of storage, a y-axis represents the width of storage, the position of equipment is represented by an (x, y) coordinate, a grid with 1 unit length is set, the grid width is set to be the same as the width of a passable road, the two-dimensional rectangular coordinate system comprises a sorting table 1, a road 2, storage robots 3, a goods shelf 4 and charging piles 5, the sorting table 1, the goods shelf 4 and the charging piles 5 are regarded as static barriers, the storage robots 3 are regarded as dynamic barriers, the storage robots move in four directions of up, down, left and right on the road 2, follow the traffic rules of right-of-way, the running directions of adjacent passable roads are different, the running directions of robots on the same road are the same, the distance between the storage robots is greater than 0.5, building an order system environment map according to the setting, and meanwhile, acquiring information of an intelligent task pool of the storage system.
Further, based on the warehouse system environment map and the total task pool, the task scheduling model is constructed, the task scheduling model comprises a task set, a robot set, path constraints, resource constraints and task priority constraints, and takes minimized path overhead, task completion time and robot energy consumption as optimization targets, and the task scheduling model construction process is as follows:
Establishing a task set S and a robot set R, wherein the task set S is a group of tasks which need to be distributed at the moment t and is expressed as S= { S i |i=1, 2,.. M }, wherein S i represents the task, i is an auxiliary variable;
Each task S i in the task set S is described in the form of Wherein the method comprises the steps ofFor the two-dimensional coordinates of the start point of the task,The two-dimensional coordinate of the termination point of the task is t i which is the task generation time, and h i which is the importance degree of the task;
Each storage robot R l in the robot set R is described in the form of Wherein the method comprises the steps ofB l is the current position coordinate of the storage robot, the electric quantity of the storage robot and Z l is the local task set of the storage robot;
Establishing a local task set Z l of the first storage robot, initializing to be empty, integrating each task into Z l after task allocation, wherein Z l comprises a plurality of tasks, namely Z l＝{sl1,sl2,…,slk, wherein s l1,sl2,…,slk is the task allocated to the first storage robot, k is the task number allocated to the first storage robot, l is an auxiliary variable, and D (r l,Zl) is the path cost of the first storage robot for completing the local task set Z l, and the calculation formula is as follows:
D(rl,Zl)＝L(sl1)+L(sl2)+...+L(slk)+M(rl,sl1)+M(sl1,sl2)+...+M(sl(k-1),slk);
Wherein L (s lk) is the distance that the first storage robot completes a single task s lk, and M (s l(k-1),slk) is the transfer distance that the first storage robot moves from the end point of the task s l(k-1) to the start point of the task s lk;
the sum of path costs of all robots completing the local task sets is represented as D all, and the calculation formula is as follows:
the task execution optimization objective is to minimize the total cost of task execution, and the specific optimization objective is as follows:
min(α*Dall+β*Twei+γ*Tmax);
Wherein alpha, beta and gamma are weight parameters, For decision variables, the value is 0 or 1,The presentation task s j is assigned to the warehousing robot r l,Indicating that task s j is not assigned to storage robot r l; The method is characterized in that the method comprises the steps of representing the number of tasks executed by a storage robot r l at a time T, T over represents the sum of delay time of all tasks, T wei represents the sum of weighted delay time of all tasks, T max represents the estimated ending time of the system for completing all tasks, and a calculation formula is as follows:
Wherein, For the time at which task s l starts to be performed by the warehousing robot, v l represents the speed of the warehousing robot r l.
Further, a task scheduling algorithm is designed based on the task scheduling model, tasks in the total task pool are distributed to all storage robots, and the task pool of each storage robot is obtained, and the task scheduling algorithm comprises the following steps:
s1, initializing, namely randomly generating an initial task scheduling scheme;
S2, neighborhood searching, namely generating a neighborhood solution of the current scheme through exchanging the sequence of tasks or adjusting the execution sequence of the tasks;
S3, evaluating the fitness, namely calculating the fitness value of the current scheme and the neighborhood solution, designing a fitness function to minimize the total cost of task execution, and specifically optimizing the target as follows:
min(α*Dall+β*Twei+γ*Tmax);
S4, an acceptance criterion is that the acceptance probability is calculated according to the adaptability values of the current solution and the neighborhood solution, whether the neighborhood solution is accepted or not is determined, the higher the temperature is, the larger the acceptance probability of the worse solution is, the lower the temperature is, and the lower the acceptance probability of the worse solution is;
s5, temperature updating, namely gradually reducing the temperature by adopting a linear decreasing strategy;
s6, terminating the condition that the maximum iteration times are reached or the temperature is reduced to a preset value.
Further, a task order optimization algorithm is designed for a task pool of each storage robot, a task execution order is optimized according to the distance between a task starting point and a task ending point and the importance of the task, and an execution plan of each storage robot is generated, and the method comprises the following steps:
s10, encoding each task sequence into an individual, wherein each individual represents a task execution sequence, each individual consists of a group of integers, and each integer represents a task;
S20, designing a fitness function, wherein the task sequence optimization aims at finding the optimal task sequence, so that the weighted sum of the running distance of the robot for completing a local task list and the weighted overtime time of all tasks is minimum, namely
Wherein, For the weight, f 1 is the estimated shortest driving distance of the storage robot to execute the local task set, f 2 is the shortest task expiration time of the existing local task set of the storage robot, and the calculation formulas of f 1 and f 2 are as follows:
Wherein, G (r l,slj) represents the accumulated travel distance of the warehousing robot r l from task s l1 to task s lj in the task list sequence, and the calculation formula is as follows:
S30, initializing a population, wherein the population comprises a plurality of individuals, each individual represents a possible task execution sequence, the size of the population is within 4N-6N, and N is the number of individual codes;
S40, selecting individuals, namely selecting individuals with high fitness from the population by using a roulette selection method as father generation;
s50, cross mutation, namely selecting two parent individuals, performing cross operation to generate new offspring individuals, and performing mutation operation on the newly generated offspring individuals;
S60, performing immune operation, namely selecting individuals with high fitness from the current population to clone to generate a plurality of cloned individuals, performing mutation operation on the cloned individuals to maintain solution diversity, and performing immune memory, namely storing high-quality solutions in an immune pool to prevent forgetting the optimal solutions;
And S70, terminating the condition that the maximum iteration times or the optimal value is not changed is reached.
Further, in the task execution process, the storage environment and the task state are monitored in real time, and when the tasks are newly added and cancelled in the total task pool, the tasks in the total task pool are redistributed based on the current task execution condition, specifically:
S100, adding tasks in a new way, namely when the number of the tasks in the total task pool reaches a preset threshold value, uniformly distributing the tasks by using a task scheduling algorithm, wherein in the process, the system only recalculates and schedules the tasks which are not distributed yet, and the distributed tasks remain unchanged;
S200, canceling the task, namely deleting the task according to the situation, if the task is not distributed to the warehousing robot, directly deleting the task, if the task is distributed but not started to be executed, deleting the task and notifying the corresponding warehousing robot to re-optimize the sequence of the residual task, and if the task is distributed and is being executed, the warehousing robot needs to return the goods shelf to the original position and then execute the residual task.
Further, for the situation of dynamic faults of the storage robots, the dispatching center rolls back the tasks which are not completed by the faulty storage robots to the dispatching center, and then task distribution is carried out again according to the current storage system environment map and the updated total task pool, and the specific steps are as follows:
s1000, task rollback, namely removing tasks which are not completed by the fault robot from a local task set and rollback to the task set;
s2000, updating the environment, namely updating an environmental map of the warehousing system to obtain the position and state change of the fault robot;
s3000, task reassigning, namely performing task assignment on the updated total task pool by utilizing a task scheduling model and a task scheduling algorithm. Compared with the prior art, the invention has the beneficial effects that:
1. The invention realizes the efficient allocation and execution sequence optimization of the tasks by designing the task scheduling algorithm and the task sequence optimization algorithm. By adopting the simulated annealing algorithm and the particle swarm hybrid genetic algorithm, task scheduling is more flexible and accurate, and the overall operation efficiency of the warehousing system is remarkably improved.
2. Aiming at dynamic events such as dynamic changes of tasks and faults of storage robots, the invention provides a corresponding rescheduling scheme, so that the task allocation can be responded and adjusted rapidly, the continuity and stability of the system under complex working conditions are effectively ensured, and the robustness and adaptability of the system are improved.
3. The invention utilizes storage resources to the maximum extent through an intelligent task scheduling and optimizing algorithm, and reduces the total cost and time of task execution. Particularly, the redistribution and optimization of incomplete tasks are beneficial to improving the utilization rate and the working efficiency of the warehousing robot, reducing the operation cost and improving the overall performance of the system and the customer satisfaction.
Drawings
FIG. 1 is a flow chart of a method of task scheduling for an intelligent warehousing system;
FIG. 2 is a warehouse system environment map;
FIG. 3 is a flow chart of a particle swarm hybrid genetic algorithm;
Fig. 4 is a task dynamic change processing procedure.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
In modern logistics and supply chain management, task scheduling of intelligent warehousing systems is becoming more and more important. The intelligent warehouse system can effectively manage and arrange warehouse tasks through an advanced scheduling algorithm and an optimization strategy, so that the efficiency and the accuracy of warehouse operation are improved. However, task scheduling systems face many challenges in complex warehousing environments, such as multiple cargo sites, multiple orders, and dynamic requirements. The intelligent warehousing system needs to maintain stable and efficient scheduling performance under various working conditions, such as processing peak orders, dynamic task changes, system interference and the like. The conventional scheduling method often has difficulty in ensuring enough scheduling precision and response speed in the face of nonlinear characteristics and external interference of the system, thereby influencing the efficiency and accuracy of warehousing operation. Therefore, the invention provides a task scheduling method of the intelligent warehousing system. In order to illustrate the feasibility of the invention, the following specific examples will be presented.
Examples
A task scheduling method of an intelligent warehousing system is adopted in the embodiment of the application, the specific implementation process is shown in figure 1, firstly, a warehousing system environment map and a total task pool of the intelligent warehousing system are obtained according to order information, the state of a warehousing robot and real-time data of warehousing environment information, a task scheduling model is constructed according to the warehousing system environment map and the total task pool, secondly, a task scheduling algorithm is designed according to the task scheduling model, tasks in the total task pool are distributed to each warehousing robot to obtain a task pool of each warehousing robot, then, a task order optimization algorithm is designed according to the task pool of each warehousing robot, task execution sequence is optimized, and finally, when abnormality is found in the task execution process, tasks in the total task pool are redistributed according to the current system state and task execution condition.
The embodiment of the application is task scheduling and optimizing based on a grid map, and the specific implementation process is as follows:
further, generating the warehousing system environment map and the total task pool according to the order information, the state of the warehousing robot and the real-time data of the warehousing environment information, wherein the warehousing system environment map is constructed by a grid method, and the specific steps comprise:
The method comprises the steps of constructing a two-dimensional rectangular coordinate system of a storage environment, wherein an x-axis represents the storage length, a y-axis represents the storage width, the position of equipment is represented by an (x, y) coordinate, a grid with 1 unit length is set, the grid width is set to be the same as the width of a passable road, as shown in fig. 2, the two-dimensional rectangular coordinate system comprises a sorting table 1, a road 2, storage robots 3, a goods shelf 4 and charging piles 5, the sorting table 1, the goods shelf 4 and the charging piles 5 are regarded as static barriers, the storage robots 3 are regarded as dynamic barriers, each storage robot moves up, down, left and right on the road 2, and follows the traffic rule of going right, the running directions of adjacent passable roads are different, the running directions of robots on the same road are the same, the keeping distance between each storage robots is greater than 0.5, according to the setting, the storage system environment map is built, order information is acquired, and an intelligent task pool of the storage system is built.
In the embodiment of the application, the technical scheme of constructing the warehouse system environment map by adopting the grid method can obviously improve the task scheduling precision and efficiency of the warehouse system. By clearly identifying the barriers and the dynamic elements, the system can realize real-time dynamic adjustment, optimize the task execution path, effectively reduce collision and traffic conflict, and improve the operation safety of the warehousing robot and the reliability of task execution.
And constructing a task scheduling model according to the warehouse system environment map and the total task pool, wherein the task scheduling model is constructed as follows:
A task set S and a robot set R are established, wherein the task set S is a group of tasks which need to be distributed at the moment t and is expressed as S= { S i |i=1, 2, & gt, m }, each task S i contains information such as a starting point, an ending point, generation time, importance degree and the like, the robot set R is a group of storage robots which can be used for executing the tasks and is expressed as R= { R i |i=1, 2, & gt, n }, and each robot R i contains information such as a current position, electric quantity, a local task set and the like;
Establishing a task set S and a robot set R, wherein the task set S is a group of tasks which need to be distributed at the moment t and is expressed as S= { S i |i=1, 2,.. M }, wherein S i represents the task, i is an auxiliary variable;
Each task S i in the task set S is described in the form of Wherein the method comprises the steps ofFor the two-dimensional coordinates of the start point of the task,The two-dimensional coordinate of the termination point of the task is t i which is the task generation time, and h i which is the importance degree of the task;
Each storage robot R l in the robot set R is described in the form of Wherein the method comprises the steps ofB l is the current position coordinate of the storage robot, the electric quantity of the storage robot and Z l is the local task set of the storage robot;
Establishing a local task set Z l of the first storage robot, initializing to be empty, integrating each task into Z l after task allocation, wherein Z l comprises a plurality of tasks, namely Z l＝{sl1,sl2,…,slk, wherein s l1,sl2,…,slk is the task allocated to the first storage robot, k is the task number allocated to the first storage robot, l is an auxiliary variable, and D (r l,Zl) is the path cost of the first storage robot for completing the local task set Z l, and the calculation formula is as follows:
D(rl,Zl)＝L(sl1)+L(sl2)+...+L(slk)+M(rl,sl1)+M(sl1,sl2)+...+M(sl(k-1),slk);
Wherein L (s lk) is the distance that the first storage robot completes a single task s lk, and M (s l(k-1),slk) is the transfer distance that the first storage robot moves from the end point of the task s l(k-1) to the start point of the task s lk;
the sum of path costs of all robots completing the local task sets is represented as D all, and the calculation formula is as follows:
the task execution optimization objective is to minimize the total cost of task execution, and the specific optimization objective is as follows:
min(α*Dall+β*Twei+γ*Tmax);
Wherein alpha, beta and gamma are weight parameters, For decision variables, the value is 0 or 1,The presentation task s j is assigned to the warehousing robot r l,Indicating that task s j is not assigned to storage robot r l; The method is characterized in that the method comprises the steps of representing the number of tasks executed by a storage robot r l at a time T, T over represents the sum of delay time of all tasks, T wei represents the sum of weighted delay time of all tasks, T max represents the estimated ending time of the system for completing all tasks, and a calculation formula is as follows:
Wherein, For the time at which task s l starts to be performed by the warehousing robot, v l represents the speed of the warehousing robot r l.
In the embodiment of the application, the task scheduling model of the intelligent warehousing system is established with obvious optimization effect. The model can improve the accuracy and efficiency of task allocation, reduce the delay time of task execution and improve the overall operation efficiency and resource utilization rate of the warehousing system. In addition, the dynamic adjustment capability of the model is beneficial to coping with the task and the change of the state of the robot, so that the adaptability and the flexibility of the system are further enhanced, the task is ensured to be completed on time, and the system performance is maximized.
Further, according to the task scheduling model, a task scheduling algorithm is designed, tasks in the total task pool are distributed to all storage robots, and the task pool of each storage robot is obtained, and the task scheduling algorithm comprises the following steps:
s1, initializing, namely randomly generating an initial task scheduling scheme;
S2, neighborhood searching, namely generating a neighborhood solution of the current scheme through exchanging the sequence of tasks or adjusting the execution sequence of the tasks;
S3, evaluating the fitness, namely calculating the fitness value of the current scheme and the neighborhood solution, designing a fitness function to minimize the total cost of task execution, and specifically optimizing the target as follows:
min(α*Dall+β*Twei+γ*Tmax);
S4, an acceptance criterion is that the acceptance probability is calculated according to the adaptability values of the current solution and the neighborhood solution, whether the neighborhood solution is accepted or not is determined, the higher the temperature is, the larger the acceptance probability of the worse solution is, the lower the temperature is, and the lower the acceptance probability of the worse solution is;
s5, temperature updating, namely gradually reducing the temperature by adopting a linear decreasing strategy;
s6, terminating the condition that the maximum iteration times are reached or the temperature is reduced to a preset value.
In the embodiment of the application, the designed task scheduling algorithm can effectively optimize the task allocation process. The algorithm remarkably improves the overall quality of the scheduling scheme through a dynamic adjustment and evaluation method, and reduces the total cost of task execution. Through optimizing the acceptance criteria and the temperature updating strategy, the algorithm can avoid falling into a local optimal solution, so that the global optimization capacity of the system and the stability of the solution are improved. The measures work together, so that the task scheduling efficiency and the system adaptability are improved.
Further, according to the task pool of each storage robot, a task order optimization algorithm is designed to optimize the task execution order, wherein the task order optimization algorithm is a particle swarm hybrid genetic algorithm, an algorithm flow chart is shown in fig. 3, and firstly, parameter setting and particle swarm are initialized from a 'start' node. Next, a time variable t=1 is set, and then the fitness calculation stage is entered. At this stage, the algorithm will calculate fitness values based on the state of the current particle. Subsequently, the individual and population optima are updated to find the best solution for each particle and the whole population, respectively. Next, an optimized exchange of individuals and groups is performed. And then judging whether the ending condition is met, if so, outputting an optimal result and ending the algorithm, and if not, increasing the time variable T by 1, and continuing the optimization of the next round. The entire process is repeated until a predetermined end condition is reached. The method comprises the following specific steps:
S10, encoding each task sequence into an individual, wherein each individual represents a task execution sequence. Each individual is made up of a set of integers, each representing a task.
S20, designing a fitness function, wherein the task sequence optimization aims at finding the optimal task sequence, so that the weighted sum of the running distance of the robot for completing a local task list and the weighted overtime time of all tasks is minimum, namely
Wherein, For the weight, f 1 is the estimated shortest travel distance of the local task set of the storage robot, f 2 is the shortest task expiration time of the existing local task set of the storage robot, and the calculation formulas of f 1 and f 2 are as follows:
Wherein, G (r l,slj) represents the accumulated travel distance of the warehousing robot r l from task s l1 to task s lj in the task list sequence, and the calculation formula is as follows:
S30, initializing a population, wherein the population comprises a plurality of individuals, each individual represents a possible task execution sequence, the size of the population is within 4N-6N, and N is the number of individual codes;
S40, selecting individuals, namely selecting individuals with high fitness from the population by using a roulette selection method as father generation;
s50, cross mutation, namely selecting two parent individuals, performing cross operation to generate new offspring individuals, and performing mutation operation on the newly generated offspring individuals;
S60, terminating the condition that the maximum iteration times or the optimal value is not changed.
In the embodiment of the application, the task order optimization method based on the particle swarm hybrid genetic algorithm remarkably improves the efficiency and the precision of task scheduling. The algorithm can optimize the task execution sequence through effective individual coding and applicability function design, and reduce the driving distance and the task over-time, thereby reducing the overall execution cost. In addition, population setting, individual selection and cross mutation operation ensure global searching capability and avoid the trouble of local optimal solution. Finally, the method improves the adaptability and the stability of the task scheduling system and enhances the overall operation efficiency of the warehousing robot.
For the situation of dynamic task change, the dynamic task change is processed by adopting an online task allocation mode, the specific processing process is shown in fig. 4, and the specific implementation process is as follows:
S100, adding tasks, namely when the number of the newly added tasks in the total task pool reaches a preset threshold value, uniformly distributing the tasks by using a task scheduling algorithm, wherein in the process, the system only recalculates and schedules the tasks which are not distributed yet, and the distributed tasks remain unchanged;
S200, task reduction, namely deleting the task according to the situation, if the task is not distributed to the warehousing robot, directly deleting the task, if the task is distributed but not started to be executed, deleting the task and notifying the corresponding warehousing robot to re-optimize the sequence of the residual task, and if the task is distributed and is being executed, the warehousing robot needs to return the goods shelf to the original position and then execute the residual task.
In the embodiment of the application, the adoption of an online task allocation mode to process the dynamically changed task significantly improves the flexibility and the response capability of the system. The method can effectively dynamically add and delete the tasks, and ensures the continuity of task scheduling and the running stability of the system. By timely adjusting and optimizing task allocation, the system can furthest reduce interruption and delay of task execution, and improves the overall operation efficiency and the resource utilization rate.
For the situation of dynamic faults of the storage robots, the dispatching center rolls back the unfinished tasks of the faulty storage robots to the dispatching center, and then the task distribution is carried out again according to the current storage system environment map and the updated total task pool, and the specific steps are as follows:
S1000, task rollback, namely removing tasks which are not completed by the fault robot from a local task set and rollback to a total task pool;
s2000, updating the environment, namely updating an environmental map of the warehousing system to obtain the position and state change of the fault robot;
s3000, task reassigning, namely performing task assignment on the updated total task pool by utilizing a task scheduling model and a task scheduling algorithm.
In the embodiment of the application, the robustness and the reliability of the system are obviously enhanced by the method for processing the dynamic faults of the storage robot. By effectively rolling back and sequencing the priorities and combining with the simulated annealing algorithm to redistribute tasks, the robot fault condition can be effectively treated, and interruption and delay of task execution are reduced. The mechanism improves the adaptability of the system to sudden faults, and ensures the smooth completion of tasks and the continuous and stable operation of the system.
The comparison between the results of the conventional particle swarm optimization algorithm and the results of the particle swarm hybrid genetic optimization algorithm is shown in the following table, wherein table 1 is the results of the particle swarm optimization algorithm, and table 2 is the results of the particle swarm hybrid genetic optimization algorithm.
Table 1 results of particle swarm optimization algorithm
Table 2 particle swarm hybrid genetic optimization algorithm
Compared with the particle swarm optimization algorithm, the particle swarm hybrid genetic optimization algorithm shows better performance under different numbers of robot configurations. Specifically, the hybrid genetic algorithm enables higher average travel distances of the AGVs, whether 20, 40 or 60 robots, with a reduction in both the total time to task completion and the average time to task completion. The particle swarm optimization algorithm is obviously improved in the aspects of path planning and task scheduling by combining a genetic algorithm, so that the robot work efficiency is higher, and the task is completed faster.
The foregoing is merely a preferred embodiment of the present invention and it should be noted that modifications and adaptations to those skilled in the art may be made without departing from the principles of the present invention, which are intended to be comprehended within the scope of the present invention.
Claims (7)
1. The task scheduling method of the intelligent warehousing system is characterized by comprising the following steps of:
s1, generating a warehousing system environment map and a total task pool according to order information, a warehousing robot state and real-time data of warehousing environment information;
s2, constructing a task scheduling model based on the warehouse system environment map and a total task pool, wherein the task scheduling model comprises task set and robot set description, path constraint, resource constraint and task priority constraint, and takes minimized path overhead, task completion time and robot energy consumption as optimization targets;
s3, designing a task scheduling algorithm based on the task scheduling model, and distributing tasks in the total task pool to each storage robot to obtain a task pool of each storage robot;
s4, designing a task sequence optimization algorithm aiming at a task pool of each storage robot, optimizing a task execution sequence according to the distance between a task starting point and a task ending point and the importance of the task, and generating an execution plan of each storage robot;
s5, monitoring the storage environment and the task state in real time in the task execution process, and reallocating the tasks in the total task pool based on the current task execution condition when the tasks are newly added and cancelled in the total task pool;
S6, for the situation of dynamic faults of the storage robots, the dispatching center rolls back the tasks which are not completed by the faulty storage robots to the dispatching center, and then task distribution is carried out again according to the current storage system environment map and the updated total task pool.
2. The task scheduling method of an intelligent warehousing system according to claim 1, wherein the warehousing system environment map and the total task pool are generated according to real-time data of order information, warehousing robot state and warehousing environment information, the specific steps of constructing the warehousing system environment map by adopting a grid method include:
The method comprises the steps of constructing a two-dimensional rectangular coordinate system of a storage environment, wherein an x-axis represents the storage length, a y-axis represents the storage width, the position of equipment is represented by an (x, y) coordinate, 1 grid of unit length is set, the grid width is set to be the same as the width of a passable road, the two-dimensional rectangular coordinate system comprises a sorting table (1), a road (2), storage robots (3), a goods shelf (4) and a charging pile (5), the sorting table (1), the goods shelf (4) and the charging pile (5) are regarded as static barriers, the storage robots (3) are regarded as dynamic barriers, each storage robot moves in four directions up, down, left and right on the road (2), the adjacent passable road is different in running direction and the same as the passable road, the distance between the storage robots is greater than 0.5, the storage system environment map is built according to the setting, and meanwhile, the total task pool of the intelligent storage system is built.
3. The task scheduling method of the intelligent warehousing system according to claim 1, wherein the task scheduling model is constructed based on the warehousing system environment map and a total task pool, the task scheduling model comprises a task set, a robot set, a path constraint, a resource constraint and a task priority constraint, and the task scheduling model is constructed by taking minimized path overhead, task completion time and robot energy consumption as optimization targets, and the task scheduling model construction process is as follows:
Establishing a task set S and a robot set R, wherein the task set S is a group of tasks which need to be distributed at the moment t and is expressed as S= { S i |i=1, 2,.. M }, wherein S i represents the task, i is an auxiliary variable;
Each task S i in the task set S is described in the form of Wherein the method comprises the steps ofFor the two-dimensional coordinates of the start point of the task,The two-dimensional coordinate of the termination point of the task is t i which is the task generation time, and h i which is the importance degree of the task;
Each storage robot R l in the robot set R is described in the form of Wherein the method comprises the steps ofB l is the current position coordinate of the storage robot, the electric quantity of the storage robot and Z l is the local task set of the storage robot;
Establishing a local task set Z l of the first storage robot, initializing to be empty, integrating each task into Z l after task allocation, wherein Z l comprises a plurality of tasks, namely Z l＝{sl1,sl2,…,slk, wherein s l1,sl2,…,slk is the task allocated to the first storage robot, k is the task number allocated to the first storage robot, D (r l,Zl) is the path cost of the first storage robot for completing the local task set Z l, and the calculation formula is as follows:
D(rl,Zl)＝L(sl1)+L(sl2)+...+L(slk)+M(rl,sl1)+M(sl1,sl2)+...
+M(sl(k-1),slk);
Wherein L (s lk) is the distance that the first storage robot completes a single task s lk, and M (s l(k-1),slk) is the transfer distance that the first storage robot moves from the end point of the task s l(k-1) to the start point of the task s lk;
the sum of path costs of all the storage robots completing the local task sets is expressed as D all, and the calculation formula is as follows:
the task execution optimization objective is to minimize the total cost of task execution, and the specific optimization objective is as follows:
min(α*Dall+β*Twei+γ*Tmax);
Wherein alpha, beta and gamma are weight parameters, For decision variables, the value is 0 or 1,The presentation task s j is assigned to the warehousing robot r l,Indicating that task s j is not assigned to storage robot r l; The method is characterized in that the method comprises the steps of representing the number of tasks executed by a storage robot r l at a time T, T over represents the sum of delay time of all tasks, T wei represents the sum of weighted delay time of all tasks, T max represents the estimated ending time of the system for completing all tasks, and a calculation formula is as follows:
Wherein, For the time at which task s l starts to be performed by the warehousing robot, v l represents the speed of the warehousing robot r l.
4. The task scheduling method of an intelligent warehousing system according to claim 1, wherein a task scheduling algorithm is designed based on the task scheduling model, tasks in a total task pool are allocated to each warehousing robot, and a task pool of each warehousing robot is obtained, and the task scheduling algorithm comprises the following steps:
s1, initializing, namely randomly generating an initial task scheduling scheme;
S2, neighborhood searching, namely generating a neighborhood solution of the current scheme through exchanging the sequence of tasks or adjusting the execution sequence of the tasks;
S3, evaluating the fitness, namely calculating the fitness value of the current scheme and the neighborhood solution, designing a fitness function to minimize the total cost of task execution, and specifically optimizing the target as follows:
min(α*Dall+β*Twei+γ*Tmax);
s4, an acceptance criterion is that the acceptance probability is calculated according to the fitness value of the current solution and the neighborhood solution and the current temperature, whether the neighborhood solution is accepted or not is determined, the higher the temperature is, the larger the acceptance probability of the worse solution is, the lower the temperature is, and the lower the acceptance probability of the worse solution is;
s5, temperature updating, namely gradually reducing the temperature by adopting a linear decreasing strategy;
s6, terminating the condition that the maximum iteration times are reached or the temperature is reduced to a preset value.
5. The task scheduling method of an intelligent warehousing system according to claim 4, wherein a task order optimization algorithm is designed for a task pool of each warehousing robot, a task execution order is optimized according to a distance between a task start point and an end point and task importance, and an execution plan of each warehousing robot is generated, comprising the steps of:
s10, encoding each task sequence into an individual, wherein each individual represents a task execution sequence, each individual consists of a group of integers, and each integer represents a task;
S20, designing a fitness function, wherein the task sequence optimization aims at finding the optimal task sequence, so that the weighted sum of the running distance of the robot for completing a local task list and the weighted overtime time of all tasks is minimum, namely
Wherein, For the weight, f 1 is the estimated shortest driving distance of the storage robot to execute the local task set, f 2 is the shortest task expiration time of the existing local task set of the storage robot, and the calculation formulas of f 1 and f 2 are as follows:
Wherein, G (r l,slj) represents the accumulated travel distance of the warehousing robot r l from task s l1 to task s lj in the task list sequence, and the calculation formula is as follows:
S30, initializing a population, wherein the population comprises a plurality of individuals, each individual represents a possible task execution sequence, the size of the population is within 4N-6N, and N is the number of individual codes;
S40, selecting individuals, namely selecting individuals with high fitness from the population by using a roulette selection method as father generation;
s50, cross mutation, namely selecting two parent individuals, performing cross operation to generate new offspring individuals, and performing mutation operation on the newly generated offspring individuals;
S60, performing immune operation, namely selecting individuals with high fitness from the current population to clone to generate a plurality of cloned individuals, performing mutation operation on the cloned individuals to maintain solution diversity, and performing immune memory, namely storing high-quality solutions in an immune pool to prevent forgetting the optimal solutions;
And S70, terminating the condition that the maximum iteration times or the optimal value is not changed is reached.
6. The method for task scheduling in an intelligent warehousing system according to claim 1, wherein the task scheduling method is characterized in that during task execution, the warehousing environment and the task state are monitored in real time, and when tasks are newly added and cancelled in a total task pool, the tasks in the total task pool are redistributed based on the current task execution condition, specifically:
S100, adding tasks in a new way, namely when the number of the tasks in the total task pool reaches a preset threshold value, uniformly distributing the tasks by using a task scheduling algorithm, wherein in the process, the system only recalculates and schedules the tasks which are not distributed yet, and the distributed tasks remain unchanged;
S200, canceling the task, namely deleting the task according to the situation, if the task is not distributed to the warehousing robot, directly deleting the task, if the task is distributed but not started to be executed, deleting the task and notifying the corresponding warehousing robot to re-optimize the sequence of the residual task, and if the task is distributed and is being executed, the warehousing robot needs to return the goods shelf to the original position and then execute the residual task.
7. The task scheduling method of an intelligent warehousing system according to claim 6, wherein for the case of a dynamic failure of the warehousing robot, the scheduling center rolls back the task which is not completed by the failed warehousing robot to the scheduling center, and then the task allocation is performed again according to the current warehousing system environment map and the updated total task pool, and the specific steps are as follows:
s1000, task rollback, namely removing tasks which are not completed by the fault robot from a local task set and rollback to the task set;
s2000, updating the environment, namely updating an environmental map of the warehousing system to obtain the position and state change of the fault robot;
s3000, task reassigning, namely performing task assignment on the updated total task pool by utilizing a task scheduling model and a task scheduling algorithm.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202510019644.3A CN119940826A (en) | 2025-01-07 | 2025-01-07 | A task scheduling method for intelligent warehousing system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202510019644.3A CN119940826A (en) | 2025-01-07 | 2025-01-07 | A task scheduling method for intelligent warehousing system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN119940826A true CN119940826A (en) | 2025-05-06 |
Family
ID=95532128
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202510019644.3A Pending CN119940826A (en) | 2025-01-07 | 2025-01-07 | A task scheduling method for intelligent warehousing system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN119940826A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN120338450A (en) * | 2025-06-19 | 2025-07-18 | 深圳市今天国际软件技术有限公司 | Delivery robot dispatching method, device, computer equipment and storage medium |
| CN120875203A (en) * | 2025-09-24 | 2025-10-31 | 国网吉林省电力有限公司营销服务中心 | Inspection robot dog scheduling method considering charging planning in electric energy meter detection scene |
-
2025
- 2025-01-07 CN CN202510019644.3A patent/CN119940826A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN120338450A (en) * | 2025-06-19 | 2025-07-18 | 深圳市今天国际软件技术有限公司 | Delivery robot dispatching method, device, computer equipment and storage medium |
| CN120875203A (en) * | 2025-09-24 | 2025-10-31 | 国网吉林省电力有限公司营销服务中心 | Inspection robot dog scheduling method considering charging planning in electric energy meter detection scene |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113128823B (en) | Automatic guided vehicle management system and method | |
| CN119940826A (en) | A task scheduling method for intelligent warehousing system | |
| Sabuncuoglu et al. | Dynamic dispatching algorithm for scheduling machines and automated guided vehicles in a flexible manufacturing system | |
| CN110264120A (en) | A kind of intelligent storage route planning system and method based on more AGV | |
| CN116523165B (en) | Collaborative optimization method for AMR path planning and production scheduling in flexible job shop | |
| CN112561194B (en) | A hybrid flow workshop production and logistics integrated scheduling method and system | |
| CN110503260A (en) | An AGV Scheduling Method Based on Dynamic Path Planning | |
| CN111027875B (en) | A multi-robot task allocation method for intelligent warehousing based on adaptive task pool | |
| CN112801484B (en) | Material distribution scheduling method and system considering batching errors | |
| CN113131554B (en) | Method for optimizing configuration of OTG wireless charging unit | |
| CN115983599B (en) | Flow shop dynamic scheduling method integrating deep reinforcement learning and multiple agents | |
| CN111400868A (en) | Distributed workshop scheduling optimization method and system with order and robot handling | |
| CN112488606A (en) | Intelligent optimization and automatic scheduling system for production logistics | |
| CN114707707A (en) | Method and system for scheduling AGV task based on improved genetic algorithm | |
| CN117930842A (en) | Multi-AGV task allocation method | |
| CN111985683A (en) | Path optimization method for material distribution of multi-target discrete assembly workshop | |
| CN118691001A (en) | Workshop scheduling method and system based on two-layer collaborative multi-objective optimization evolution | |
| CN116859833A (en) | AGV dynamic scheduling and path planning method based on tabu search | |
| Schwab | A decentralized control strategy for high density material flow systems with automated guided vehicles | |
| Ge et al. | Research on online scheduling method for flexible assembly workshop of multi-agv system based on assembly island mode | |
| CN119472742A (en) | A method for allocating search and rescue tasks for UAV swarm based on discrete particle swarm | |
| CN118605513A (en) | Parallel optimization method for robot path planning based on load balancing and multi-search strategy | |
| Song et al. | A dynamic task assignment optimization method for multi-AGV system based on genetic algorithm | |
| CN119476872B (en) | Task scheduling method for stereoscopic warehouse apparatus, device and medium | |
| CN120471252B (en) | Construction robot operation path optimization method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |