CN116339955B - Local optimization method, device and computer equipment for computing-for-communication framework - Google Patents

Abstract

The application relates to a local optimization method, a device and computer equipment for a computing communication framework. The method comprises the following steps: and constructing a scoring function of tasks related to execution positions of all the agents, calculating tasks to be executed by adopting a local optimization method according to the scoring function to obtain conflict-free scheduling, calculating the tasks to be executed by adopting the local optimization method for all the agents to obtain conflict-free task scheduling, and transmitting the conflict-free task scheduling to other agents in a communication mode if the existence of conflict is detected when global convergence confirmation is carried out, sending a schedule to the other agents, and eliminating task conflict generated in a global convergence stage by adopting a consistency elimination principle to obtain a network multi-agent scheduling scheme. The application can reduce the communication times and improve the dispatching generation speed.

Description

Local optimization method, device and computer equipment for computing communication framework

Technical Field

The present application relates to the field of intelligent agent scheduling technology, and in particular to a local optimization method, device and computer equipment for a computing-for-communication framework.

Background Art

Multi-agent scheduling is nothing new, but the swarm craze in recent years has reactivated it. As an important application form of multi-agent theory, the swarm gives multi-agents new characteristics such as flexibility, locality and autonomy, making it very valuable in engineering. It can be expressed as a group of vehicles, drones, robots or other agents with individual dynamics. Depending on the individual cost and function, the swarm can adopt group behavior (low cost, single function) or distributed precise scheduling (high cost, complete function). For the former, the swarm algorithm based on threshold response can realize aggregation, dispersion and other behaviors. For the latter, a multi-agent scheduling algorithm is needed to achieve precise coordination such as assembly and scheduling.

Although scheduling is still a process of matching "agent-task" pairs with time series. The swarm is mostly composed of low-cost agents, which puts higher requirements on computing, communication and storage, which need to be considered in practical applications. Among these influencing factors, due to the rapid development of chips, miniaturized, low-power, and high-performance computing units are emerging in an endless stream, such as NVIDIA, AMD or Intel series products, the contradiction is not so prominent. Although communication technologies such as self-organizing networks have made great progress, it is still a more challenging factor due to environmental influences, especially in complex terrain occlusion or electromagnetic environments. The communication between agents is inevitably affected by the environment, resulting in varying degrees of errors, packet loss, delays and even interruptions, which affects the generation time of the schedule and even the success or failure. Therefore, how to reduce the reliance on risk communication and enhance the robustness and timeliness of schedule generation is a problem that needs to be considered.

The traditional approach is to solve the scheduling problem through two stages: calculation and communication. In the calculation stage, the schedule is obtained by local agent calculation, and in the communication stage, conflicts are resolved by communicating with adjacent agents about their arrangements. The two processes are iterated to obtain a scheduling solution. As far as the two stages are concerned, the former stage focuses on improving scheduling efficiency, while the latter stage focuses on improving optimization performance. Therefore, making the total scheduling generation time short and the communication time short become issues that need to be urgently addressed.

Summary of the invention

Based on this, it is necessary to provide a network multi-agent scheduling computing and communication system to address the above technical problems.

A local optimization method for a computation-for-communication framework, the method comprising:

Construct scoring functions for each agent to perform position-related tasks;

According to the scoring function, a local optimization method is used to calculate the tasks to be executed to obtain a conflict-free schedule; wherein, in the local optimization method, each agent interacts with each other through communication, and the location information of the agent is determined by the score of the scoring function corresponding to the task executed by other agents stored in the agent, and the local agent included in the local optimization method is selected through a preset task-related selection strategy;

Calculate the tasks to be executed by each agent through a local optimization method to obtain a conflict-free task schedule. When confirming the global convergence, if a conflict is detected, the conflict-free task schedule is transmitted to other agents through communication;

The schedule is sent to other agents, and the consistency elimination principle is used to eliminate the task conflicts generated in the global convergence stage to obtain the network multi-agent scheduling plan.

In one embodiment, the scoring function is expressed as:

;

in, It is an intelligent agent Execute the task The score, Representing an Agent timetable, Representing an Agent According to the schedule Execute the task The return, It's for the mission The discount factor for the return, For intelligent agents Along the schedule Or the agent's location arrival task Estimated time of location, For the task Initial mission reward.

In one embodiment, the optimization problem for the agent to perform task scheduling is:

;

Where n represents the number of agents, m represents the number of tasks, is a binary variable, when When the intelligent agent Execute the task ,when , indicating that the agent Not performing tasks , It is an intelligent agent Execute the task The constraints of the optimization problem are:

;

;

;

;

in, Representing an Agent In task scheduling Execute tasks in time.

In one of the embodiments, the local optimization includes: task exchange and local sampling;

Increasing the local scope score by the task exchange; the local scope score refers to the score of all agents performing tasks in the local scope;

The task scores of other agents in the local range are obtained through the local sampling, so as to perform local estimation and obtain conflict-free task scheduling.

In one embodiment, the task exchange process includes:

Set the task exchange strategy to:

;

;

;

in, It is an intelligent agent Select Exchange strategy, local score It is an intelligent agent With Agent The sum of the scores of Is to make the intelligent agent Mission With Agent Mission exchange, is the local score after the exchange.

In one embodiment, the local sampling process is:

According to the current schedule, task information, location information and scoring function of all agents in the local range, the local auction is constructed as:

;

in, represents the local estimation function of agent i for other agents in the local range, The local auction function of other agents k at the current time t, Indicates the task Task information, Represents the location information of agent k.

In one of the embodiments, according to the scoring function, distance information between the agent and the task to be performed is calculated;

The location information of the intelligent agent is determined based on at least three of the distance information.

In one embodiment, the displayed agent is the current agent that solves the winner vectors of other agents in the current iteration from the local memory, and solves other agents that have task conflicts with the current agent; the implicit agent is an agent that may conflict when there is no displayed task conflict.

A local optimization device for a computing-for-communication framework, the device comprising:

The scoring function building module is used to build the scoring function for each agent to perform position-related tasks;

A local optimization module, used to calculate the tasks to be executed using a local optimization method according to the scoring function to obtain a conflict-free schedule; wherein, in the local optimization method, each agent interacts with each other through communication, determines the location information of the agent through the scores of the scoring function corresponding to the tasks executed by other agents stored in the agent, and selects the local agent included in the local optimization method through a preset task-related selection strategy;

The scheduling module is used to calculate the tasks to be executed by each intelligent agent through a local optimization method to obtain a conflict-free task schedule. When confirming the global convergence, if a conflict is detected, the conflict-free task schedule is transmitted to other intelligent agents through communication;

The solving module is used to send schedules to other agents and adopt the consistency elimination principle to eliminate the task conflicts generated in the global convergence stage to obtain the network multi-agent scheduling solution.

In one embodiment, a computer device includes a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of:

Construct scoring functions for each agent to perform position-related tasks;

According to the scoring function, a local optimization method is used to calculate the tasks to be executed to obtain a conflict-free schedule; wherein, in the local optimization method, each agent interacts with each other through communication, and the location information of the agent is determined by the score of the scoring function corresponding to the task executed by other agents stored in the agent, and the local agent included in the local optimization method is selected through a preset task-related selection strategy;

Calculate the tasks to be executed by each agent through a local optimization method to obtain a conflict-free task schedule. When confirming the global convergence, if a conflict is detected, the conflict-free task schedule is transmitted to other agents through communication;

The schedule is sent to other agents, and the consistency elimination principle is used to eliminate the task conflicts generated in the global convergence stage to obtain the network multi-agent scheduling plan.

The local optimization method, device and computer equipment of the above computation-for-communication framework, under the idea of computation-for-communication, reduce the number of communications by appropriately increasing the amount of computation, thereby maintaining or even reducing the scheduling generation time. A new scoring function is designed, which is more optimized and can provide support for inferring the positions of other agents. Under this, a local optimization method is proposed to further improve the optimization and timeliness of scheduling. Task exchange and task sampling methods are used to explore better scheduling methods to avoid local optimal problems caused by individual greed. Conflicts between agents are resolved in advance through local estimation to reduce the number of iterations. At the same time, the proposed task-related agent selection strategy is adopted to spend the computation on key agents.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG1 is a comparison diagram of the differences between a communication computing framework and a traditional market framework in one embodiment, wherein (a) is a schematic diagram of a traditional market-based framework, and (b) is a schematic diagram of the communication computing framework of the present application;

FIG2 is a schematic flow chart of a local optimization method for computing a communication framework in one embodiment;

FIG3 is a framework diagram of a local optimization method for computing a communication framework in one embodiment;

FIG. 4 is a schematic diagram of the interior of a computer device in one embodiment.

DETAILED DESCRIPTION

In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

As shown in Figure 1, Figure 1 (a) is a schematic diagram of a traditional market-based framework, and Figure 1 (b) is a schematic diagram of a communication computing framework of the present application. It can be seen from the traditional market-based framework of Figure 1 (a) that the total scheduling time includes planning time, communication time and waiting time, and the planning time, communication time and waiting time include computing time and communication time. In the computing-for-communication framework of the present invention, due to the introduction of the computing-for-communication conflict resolution strategy, although the overall incremental planning time increases, the communication time is greatly reduced and the scheduling efficiency is improved.

In one embodiment, as shown in FIG2 , a local optimization method of a computation-for-communication framework is provided, as follows:

Step 202, construct a scoring function for each agent to perform location-related tasks.

The scoring function is the basis for the agent to select tasks and drives the generation of the agent's schedule.

Step 204 , according to the scoring function, a local optimization method is used to calculate the tasks to be executed to obtain a conflict-free schedule.

In the local optimization method, each agent interacts with each other through communication. The location information of the agent is determined by the scores of the scoring functions corresponding to the tasks stored by other agents in the agent's storage, and the local agents included in the local optimization method are selected through the pre-set task-related selection strategy.

In this step, although the amount of calculation is increased by local optimization, the number of communications can be greatly reduced. With the great expansion of computing resources, the calculation-for-communication framework of the present invention can ensure that the scheduling generation time is short and the number of communications can be reduced.

Step 206, calculate the tasks to be executed for each intelligent agent through a local optimization method to obtain a conflict-free task schedule. When confirming the global convergence, if a conflict is detected, the conflict-free task schedule is transmitted to other intelligent agents through communication.

Through local optimization, the intention is to eliminate conflicts between tasks, so as to obtain conflict-free task scheduling, and then detect global convergence.

Step 208, sending a schedule to other agents, using the consistency elimination principle to eliminate task conflicts generated in the global convergence phase, and obtaining a network multi-agent scheduling solution.

In the local optimization method of the computation-for-communication framework mentioned above, under the idea of computation-for-communication, the number of communications is reduced by appropriately increasing the amount of computation, thereby maintaining or even reducing the scheduling generation time. A new scoring function is designed, which is more optimized and can provide support for inferring the positions of other agents. Under this, a local optimization method is proposed to further improve the optimization and timeliness of scheduling. Task exchange and task sampling methods are used to explore better scheduling methods to avoid local optimal problems caused by individual greed. Conflicts between agents are resolved in advance through local estimation to reduce the number of iterations. At the same time, the proposed task-related agent selection strategy is adopted to spend the computation on key agents.

The scheduling calculation of multi-agents in the network is generally divided into two stages, namely the calculation stage and the communication stage. In the calculation stage, the agents all pursue the common goal of maximizing the score, but due to the limited capabilities of the agents, their scores have upper bounds. In the communication stage, each agent follows the consistency rule of the highest bidder, that is, the task belongs to the agent with the highest score, and the task conflict can be resolved by the score. Therefore, as long as there is no deadlock problem caused by task score oscillation, the market-based method will definitely converge. There are three main reasons that affect the convergence speed: the first is the number of task conflicts, which is related to the distribution of agents or tasks and the performance of the algorithm. For example, some greedy algorithms are prone to task conflicts; the second is the monotonicity of task scores. If the score is not monotonic, the task preferred by the agent may be lost due to low scores, and tasks need to be reselected for iteration, which may even cause the task selection cycle to deadlock; the third is the update of task scores. After deleting or adding tasks, once the task score is updated, the attributes of the task may change, and it is necessary to iterate again to resolve task conflicts.

In summary, the key to improving the scheduling efficiency of multi-agent networks lies in how to accelerate convergence, that is, reduce the number of iterations. Therefore, in the present invention, the task conflict is reduced by increasing the computational workload of other agents, thereby reducing iterations. Although the computational time is prolonged, the number of iterations is reduced, thereby reducing the communication rounds. The increase in computational workload is mainly due to local optimization including other agents, which is expected to reduce the number of communications and keep the total scheduling generation time unchanged or even reduce it.

In one of the embodiments, as shown in FIG3 , the computing-for-communication system of the present invention, similar to traditional computing, retains two iterative stages. In the computing stage, the agent eliminates tasks that do not belong to it according to the task attributes after conflict resolution, and then calculates conflict-free tasks through local optimization and adds them to the agent schedule. Next, convergence confirmation is performed. If there is a conflict, the schedule is passed to other agents to resolve the conflict in the communication stage, otherwise the final schedule will be obtained. In the communication stage, the agent sends its own schedule to other agents in the form of task winners, task winning bids and timestamps, and receives the schedules of other agents, and then resolves task conflicts according to consistency rules. The commonly used consistency rule is that the highest bidder wins, that is, the agent with the highest task bid bids higher than other agents to obtain the task.

In one embodiment, the scoring function is expressed as:

;

in, It is an intelligent agent Execute the task The score, Representing an Agent timetable, Representing an Agent According to the schedule Execute the task The return, It's for the mission The discount factor for the return, For intelligent agents Along the schedule Or the agent's location arrival task Estimated time of location, For the task Initial mission reward.

In one embodiment, the optimization problem for the agent to perform task scheduling is:

;

Where n represents the number of agents, m represents the number of tasks, is a binary variable, when When the intelligent agent Execute the task ,when , indicating that the agent Not performing tasks , the constraints of the optimization problem are:

;

;

;

;

in, Representing an Agent In task scheduling Execute tasks in time.

Specifically, multi-agent scheduling is n agents assigning m tasks to form their own schedules. The scheduling of pi is to assign the execution order and time of tasks to meet the constraints such as the start time and deadline of the tasks. It can generally be expressed as a constrained optimization problem. Improvements were made to introduce tasks For Agents The schedule of drives the agent to explore new tasks based on the previous task, regardless of whether the new task is near other agents, which may lead to blind outward exploration. Different from the common scoring function, the latter term is introduced to drive the agent to select tasks near its location, so that the agent tries to avoid competing for tasks near other agents, which helps to reduce conflicts between agents. Each agent tries to choose tasks that are closer to itself, which can avoid agents with closer tasks waiting for other agents far away to perform time-limited tasks and cause task failure. In addition, the introduction of It also implies the distance information between the agent and the task.

In one embodiment, local optimization includes: task exchange and local sampling, where the local scope score is increased through task exchange; the local scope score refers to the score of all agents performing tasks within the local scope; and the task scores of other agents within the local scope are obtained through local sampling, thereby performing local estimation to obtain conflict-free task scheduling.

Specifically, local optimization means that the agent considers the scheduling of other agents in the local scope, rather than only considering its own scheduling like individual optimization. This allows the agent to avoid possible conflicts with other agents in the local scope in advance and try to escape the local optimum that the individual optimization falls into. Local optimization mainly includes two steps: the first step is task exchange, which improves the performance of scheduling by exchanging assigned tasks between agents; the second is local sampling, which samples various feasible schedules, selects the optimal schedule, and further improves the scheduling performance. In these two steps, task conflicts will be resolved due to the calculation of the schedules of other agents.

In one embodiment, the task exchange process includes:

Set the task exchange strategy to:

;

;

;

in, It is an intelligent agent Select Exchange strategy, local score It is an intelligent agent With Agent The sum of the scores of Is to make the intelligent agent Mission With Agent Mission exchange, is the local score after the exchange.

Specifically, in the task exchange step, the local scope is the two agents that exchange tasks, and in local sampling, the local scope is multiple agents related to the task. Therefore, the performance of local optimization is evaluated by the local score, which is the sum of the scores of all agents in the specified local scope. Local score for:

;

in, are all agents in the local scope, which are different during the task exchange and local sampling phases.

For the task exchange process, the agent Try it with other agents Swap tasks to improve local optimization, which can be reflected by the local score of the two agents. Since there may be many feasible swaps, the optimal swap is chosen that increases the local score the most.

Specifically, since task exchange requires a large number of traversals, a heuristic algorithm can be used to solve the above task exchange strategy to improve computational efficiency.

In one embodiment, the local sampling process is:

According to the current schedule, task information, location information and scoring function of all agents in the local range, the local auction is constructed as:

;

in, represents the local estimation function of agent i for other agents in the local range, The local auction function of other agents k at the current time t, Indicates the task Task information, Represents the location information of agent k.

Specifically, in each sampling, the agent Calculate all relevant agents The task scores of the agents are estimated, and their task selection and result progress are estimated, so as to avoid task conflicts with other agents in advance and reduce the communication required between agents. This process of estimating the scheduling of other agents is called local estimation. , mainly based on all agents in the local area The current schedule of all tasks Information , Agent Location and the scoring function To construct a local auction.

Specifically, the specific steps of local sampling are as follows:

1. Mission failure

Mission Failure Mainly prevents agents Select Task Optimal Unqualified Task is the task of maximizing the estimated local score. If If it is 0, there are no unrestricted tasks:

;

;

2. Unqualified candidates for tasks

The unqualified tasks consist of optional unqualified tasks, which are all the agents in the local scope. Related tasks :

;

Among them, the intelligent Explicit task candidates Consists of tasks that outbid the opponent, implicit task candidates Consists of potentially conflicting tasks:

;

;

The following example shows that the schedule of Agent-1 and Agent-2 is shown in State-1. Through task exchange, the agents get a better schedule (State-2). In State-2, Agent-3 obtains two samples through local sampling, where the agent of Sampling-1 knows in advance that Agent-2 will obtain the task through estimation, giving up Task-G and obtaining Task-H. In Sampling-2, Agent-3 actively disqualifies Task-H and gives way to Agent-2, choosing Task-I which is farther away and has a higher local score.

In one embodiment, in order to implement the above-mentioned local optimization process, it is also necessary to infer the positions of other agents. These positions are usually obtained through communication. However, since the communication protocol of the traditional method does not include the position of the agent, it is necessary to add additional communication protocols and traffic. In order to maintain the low traffic advantage of the traditional method, the present invention attempts to infer the positions of other agents from the historical data received by the communication without adding new communications.

In the present invention, the intelligent agent To the task The bid of the agent is calculated by the score. To the task The estimated arrival time of the agent With the task The direct or cumulative distance between , so the agent To the task The distance can be obtained by the inverse function of the scoring function:

;

Among them, all agents know the task Location ,value and discount factor , Agent Cruising speed Known.

Computational Agents To the task Distance Afterwards, the agent The location is based on the task The position is the center of the circle, the distance From the three circles to determine an intersection, it can be seen that the intelligent agent is obtained The location only requires three different tasks .

In another embodiment, since the calculation of too many other agents will greatly increase the calculation time, an appropriately increased amount of calculation should be used for "competitive" agents that may compete for tasks. The core idea of selecting these competitive agents is task relevance. Therefore, it is also necessary to specify a corresponding task-related agent selection strategy.

1. Task-related Agents

Task-Related Agents Interested in the same task, there may be conflicts, by explicit agents and implicit agents composition:

;

2. Display Agent

An explicit agent is an agent that currently has an explicit task conflict. From local memory Find the current iteration Other Agents The winner vector , and then find all the agents Scheduling Agents with Task Conflict :

;

3. Implicit Agents

Implicit agents are agents that may conflict in the future when there is no explicit task conflict at present. Various strategies for implicit agents can be considered. , mainly using other agents The winner's bid To infer the timetable or Location :

;

Explore the implicit agent's strategy in a limited time. From the agent's current schedule Start by exploring the tasks in the schedule one by one. Centered on the mission The cumulative distance The closest potential agent within a circle of radius:

;

It should be understood that, although the various steps in the flowchart of Fig. 2 are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps is not strictly limited in order, and these steps can be executed in other orders. Moreover, at least a part of the steps in Fig. 2 may include a plurality of sub-steps or a plurality of stages, and these sub-steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these sub-steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the sub-steps or stages of other steps.

In one embodiment, as shown in FIG3 , a local optimization device for a computation-for-communication framework is provided, comprising:

A scoring function construction module 302 is used to construct a scoring function for each agent to perform a position-related task;

The local optimization module 304 is used to calculate the tasks to be executed according to the scoring function using a local optimization method to obtain a conflict-free schedule; wherein, in the local optimization method, each agent interacts with each other through communication, determines the location information of the agent through the scores of the scoring function corresponding to the tasks executed by other agents stored in the agent, and selects the local agent included in the local optimization method through a preset task-related selection strategy;

The scheduling module 306 is used to calculate the tasks to be executed by each agent through a local optimization method to obtain a conflict-free task schedule. When confirming the global convergence, if a conflict is detected, the conflict-free task schedule is transmitted to other agents through a communication method.

The solution module 308 is used to send a schedule to other agents, and use the consistency elimination principle to eliminate the task conflicts generated in the global convergence stage to obtain a network multi-agent scheduling solution.

In one embodiment, the scoring function is expressed as:

;

in, It is an intelligent agent Execute the task The score, Representing an Agent timetable, Representing an Agent According to the schedule Execute the task The return, It's for the mission The discount factor for the return, For intelligent agents Along the schedule Or the agent's location arrival task Estimated time of location, For the task Initial mission reward.

In one embodiment, the optimization problem for the agent to perform task scheduling is:

;

Where n represents the number of agents, m represents the number of tasks, is a binary variable, when When the intelligent agent Execute the task ,when , indicating that the agent Not performing tasks , It is an intelligent agent Execute the task The constraints of the optimization problem are:

;

;

;

;

in, Representing an Agent In task scheduling Execute tasks in time.

In one of the embodiments, the local optimization includes: task exchange and local sampling;

Increasing the local scope score by the task exchange; the local scope score refers to the score of all agents performing tasks in the local scope;

The task scores of other agents in the local range are obtained through the local sampling, so as to perform local estimation and obtain conflict-free task scheduling.

In one embodiment, the task exchange process includes:

Set the task exchange strategy to:

;

;

;

in, It is an intelligent agent Select Exchange strategy, local score It is an intelligent agent With Agent The sum of the scores of Is to make the intelligent agent Mission With Agent Mission exchange, is the local score after the exchange.

In one embodiment, the local sampling process is:

According to the current schedule, task information, location information and scoring function of all agents in the local range, the local auction is constructed as:

;

in, represents the local estimation function of agent i for other agents in the local range, The local auction function of other agents k at the current time t, Indicates the task Task information, Represents the location information of agent k.

In one embodiment, the determining the location information of the agent includes: calculating the distance information between the agent and the task to be performed according to the scoring function; and determining the location information of the agent according to at least three of the distance information.

In one embodiment, the task-related selection strategy includes: an explicit agent and an implicit agent; the explicit agent is the current agent that solves the winner vector of other agents in the current iteration from the local memory, and solves other agents that have task conflicts with the current agent; the implicit agent is an agent that may conflict when there is no explicit task conflict.

In one embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be shown in FIG4. The computer device includes a processor, a memory, a network interface, a display screen, and an input device connected via a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor, a local optimization method for a computing-communication framework is implemented. The display screen of the computer device may be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device may be a touch layer covered on the display screen, or a key, trackball or touchpad provided on the housing of the computer device, or an external keyboard, touchpad or mouse, etc.

Those skilled in the art will understand that the structure shown in FIG. 4 is merely a block diagram of a partial structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

In one embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the method in the above embodiment when executing the computer program.

In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps of the method in the above embodiment are implemented.

Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application may include non-volatile and/or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

The technical features of the above embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

Claims (8)

1. A local optimization method for a computation-for-communication framework, characterized in that the method comprises: Construct scoring functions for each agent to perform position-related tasks; According to the scoring function, a local optimization method is used to calculate the tasks to be executed to obtain a conflict-free schedule; wherein, in the local optimization method, each agent interacts with each other through communication, and the location information of the agent is determined by the score of the scoring function corresponding to the task executed by other agents stored in the agent, and the local agent included in the local optimization method is selected through a preset task-related selection strategy; Calculate the tasks to be executed by each agent through a local optimization method to obtain a conflict-free task schedule. When confirming the global convergence, if a conflict is detected, the conflict-free task schedule is transmitted to other agents through communication; Send the schedule to other agents, use the consistency elimination principle to eliminate the task conflicts generated in the global convergence phase, and obtain the network multi-agent scheduling plan; The scoring function is expressed as: ; in, It is an intelligent agent Execute the task The score, Representing an Agent timetable, Representing an Agent According to the schedule Execute the task The return, It's for the mission The discount factor for the return, For intelligent agents Along the schedule Or the agent's location arrival task Estimated time of location, For the task Initial mission reward; The task-related selection strategies include: explicit agents and implicit agents; The display agent is the current agent that solves the winner vectors of other agents in the current iteration from the local memory, and solves other agents that have task conflicts with the current agent; The implicit agent is an agent that may conflict when there is currently no explicit task conflict. 2. The local optimization method of the computation-for-communication framework according to claim 1 is characterized in that the optimization problem of task scheduling performed by the agent is: ; Where n represents the number of agents, m represents the number of tasks, is a binary variable, when When the intelligent agent Execute the task ,when , indicating that the agent Not performing tasks , It is an intelligent agent Execute the task The constraints of the optimization problem are: ; ; ; ; in, Representing an Agent In task scheduling Execute tasks in time. 3. The local optimization method of the computation-for-communication framework according to claim 1, characterized in that the local optimization includes: task exchange and local sampling; Increasing the local scope score by the task exchange; the local scope score refers to the score of all agents performing tasks in the local scope; The task scores of other agents in the local range are obtained through the local sampling, so as to perform local estimation and obtain conflict-free task scheduling. 4. The local optimization method of the computation-for-communication framework according to claim 3, characterized in that the task exchange process comprises: Set the task exchange strategy to: ; ; ; in, It is an intelligent agent Select Exchange strategy, local score It is an intelligent agent With Agent The sum of the scores of Is to make the intelligent agent Mission With Agent Mission exchange, is the local score after the exchange. 5. The local optimization method of the computation-for-communication framework according to claim 3, characterized in that the local sampling process is: According to the current schedule, task information, location information and scoring function of all agents in the local range, the local auction is constructed as: ; in, represents the local estimation function of agent i for other agents in the local range, The local auction function of other agents k at the current time t, Indicates the task Task information, Represents the location information of agent k. 6. The local optimization method of the computation-for-communication framework according to claim 1, characterized in that the determining the location information of the intelligent agent comprises: According to the scoring function, calculating the distance information between the agent and the task to be performed; The location information of the intelligent agent is determined based on at least three of the distance information. 7. A local optimization device for a computation-for-communication framework, characterized in that the device comprises: The scoring function building module is used to build the scoring function for each agent to perform position-related tasks; A local optimization module, used to calculate the tasks to be executed using a local optimization method according to the scoring function to obtain a conflict-free schedule; wherein, in the local optimization method, each agent interacts with each other through communication, determines the location information of the agent through the scores of the scoring function corresponding to the tasks executed by other agents stored in the agent, and selects the local agent included in the local optimization method through a preset task-related selection strategy; The scheduling module is used to calculate the tasks to be executed by each intelligent agent through a local optimization method to obtain a conflict-free task schedule. When confirming the global convergence, if a conflict is detected, the conflict-free task schedule is transmitted to other intelligent agents through a communication method. The solution module is used to send schedules to other agents, and use the consistency elimination principle to eliminate the task conflicts generated in the global convergence stage to obtain the network multi-agent scheduling solution; The scoring function is expressed as: ; in, It is an intelligent agent Execute the task The score, Representing an Agent timetable, Representing an Agent According to the schedule Execute the task The return, It's for the mission The discount factor for the return, For intelligent agents Along the schedule Or the agent's location arrival task Estimated time of location, For the task Initial mission reward; The task-related selection strategies include: explicit agents and implicit agents; The display agent is the current agent that solves the winner vectors of other agents in the current iteration from the local memory, and solves other agents that have task conflicts with the current agent; The implicit agent is an agent that may conflict when there is currently no explicit task conflict. 8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein the processor implements the steps of the method according to any one of claims 1 to 6 when executing the computer program.