Multi-agent transportation task matching method and system

Abstract

The invention discloses a multi-agent transportation task matching method and system, and relates to the agent cooperation technology, and the method comprises the steps: obtaining the state information and task demand information of agents, and initializing an efficiency matrix; performing task matching by using a digital Hungary algorithm; after task matching, standardizing the efficiency matrix; and updating the state of the unmanned aerial vehicle and the task state, and repeating task matching. According to the embodiment of the invention, the task initial weight and overtime penalty are considered when the efficiency matrix is constructed, and the Hungary algorithm is digitalized; and meanwhile, through a loop iteration mode, the efficiency matrix is dynamically updated, so that dynamic automatic matching of transportation tasks is realized, and efficient completion of tasks such as emergency rescue and material transportation is guaranteed.

Description

Technical Field

[0001] This application relates to the field of intelligent agent collaboration technology, and in particular to a method and system for matching multi-agent transportation tasks. Background Technology

[0002] Intelligent agents such as drones, unmanned vehicles, and unmanned dogs are particularly suitable for tasks such as material transportation and rescue in specific scenarios due to their strong environmental adaptability.

[0003] However, the current variety of intelligent agents, such as drones, unmanned vehicles, and unmanned dogs, is vast and fragmented, making it difficult to form a cohesive rescue force. Furthermore, common task matching methods are based on static assumptions, directly matching tasks with intelligent agents in a single step, without considering the varying urgency and time constraints of actual task execution. Therefore, there is an urgent need to develop an automatic multi-agent transportation task matching method to achieve dynamic and automatic task matching.

[0004] Chinese patent "CN114037387A Method and Equipment for Intelligent Matching of Available Transportation Capacity Resources". Addressing the transportation capacity scheduling problem in the express delivery industry, this patent provides a step-by-step, semi-automated method, improving upon the current situation where cargo volume and transportation capacity scheduling mainly rely on manual operation. The main steps of this patent are as follows: The first step is to manually enter the vehicle request (originating center, destination station, time, etc.); The second step is for the system to query the database for available vehicles based on the request; The third step is to classify vehicles according to their status (such as vehicles available in the parking lot, vehicles en route without a plan, etc.). The fourth step is for the system to display the categorized available vehicles to the dispatcher.

[0005] However, this patent still requires dispatchers to make manual decisions by displaying information, without considering the urgency of the task or realizing automated decision-making.

[0006] Traditional task allocation methods include exhaustive search, planning methods, simulated annealing, tabu search, neural networks, genetic algorithms, ant colony optimization, and the Hungarian algorithm. Among these, exhaustive search suffers from exponentially increasing computational costs for large-scale problems, making it almost impractical. Planning methods rely on specialized mathematical knowledge for model construction and solution, potentially leading to lengthy solutions for complex problems. Simulated annealing has a relatively complex computational process, with parameter settings affecting the results and not always guaranteeing a globally optimal solution. Tabu search may miss optimal solutions due to improper tabu list settings, limiting the exploration of new solutions. Neural network methods require substantial data and computational resources for training, resulting in a complex training process and poor interpretability. Genetic algorithms have complex parameter settings, high computational costs, and potentially slow convergence. Ant colony optimization has slow initial convergence, and parameter settings affect algorithm performance, making it unsuitable for emergency search and rescue scenarios. Summary of the Invention

[0007] This application provides a multi-agent transportation task matching method and system to achieve dynamic and automatic matching of transportation tasks, ensuring the efficient completion of tasks such as emergency rescue and material transportation.

[0008] This application provides a multi-agent transportation task matching method, including: Obtain the agent's state information and task requirement information, and initialize the efficiency matrix; A digital Hungarian algorithm is used for task matching; After task matching, the efficiency matrix is ​​standardized; Update the drone status and mission status, and repeat the mission matching process.

[0009] This application provides a multi-agent transportation task matching system, including a processor and a memory. The memory stores a computer program, which, when executed by the processor, implements the steps of the aforementioned multi-agent transportation task matching method.

[0010] In this embodiment, the initial weight of the task and the timeout penalty are considered when constructing the efficiency matrix, and the Hungarian algorithm is digitized. At the same time, the efficiency matrix is ​​dynamically updated through a cyclic iteration mode, thereby realizing the dynamic automatic matching of transportation tasks and ensuring the efficient completion of tasks such as emergency rescue and material transportation.

[0011] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description

[0012] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the scope of this application. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 This is a schematic diagram of the overall process of the multi-agent transportation task matching method in an embodiment of this application; Figure 2 This is a schematic diagram of the task matching process of the digital Hungarian algorithm for the multi-agent transportation task matching method in this application embodiment. Detailed Implementation

[0013] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0014] This application provides a multi-agent transportation task matching method, such as... Figure 1 As shown, it includes: In step S101, the state information and task requirement information of the agent are obtained, and the efficiency matrix is ​​initialized. This embodiment uses the parameter representation of the standard Hungarian algorithm: i is the task number, j is the agent, and n is the maximum value of i and j. It is the time required for agent i to complete task j. These are the decision variables for the Hungarian algorithm.

[0015] Based on this, the embodiments of this application introduce a state function for agent j. Its value range is 0 (idle), 1 (in task), 2 (fault); (Introduction) This represents the time when agent j was most recently assigned a task; (Introduction) This indicates the time required for drone j to execute the current task i after task assignment.

[0016] Introduce base coordinates B; introduce current time. Time step Introduce the urgency parameter for task i. Introducing initial weights ,in It is the initial weight of the task points, reflecting the task type (e.g., a value of 8 for serious injury rescue points and 3 for minor injury rescue points). The weight decay coefficient (e.g., 0.5 for serious injury rescue points and 0.1 for minor injury rescue points) reflects the urgency of the task; a penalty coefficient is introduced for task i to time out. , .

[0017] Improved time efficiency matrix: in, The specific purpose of introducing a timeout penalty function is as follows: When the time required for agent j to execute task i is less than the required completion time for task i. , At this point, the time cost remains unchanged.

[0018] When the time required for agent j to execute task i is greater than the required completion time for task i... , At this point, the time cost will increase, which will be reflected in the penalty for exceeding the time limit.

[0019] make The time efficiency matrix is ​​initialized as follows: .

[0020] In step S102, a digital Hungarian algorithm is used for task matching; In step S103, after task matching, the efficiency matrix is ​​standardized; In step S104, the drone status and mission status are updated, and mission matching is repeated.

[0021] In some embodiments, using a digitized Hungarian algorithm for task matching includes: The objective function constructed is: in, Represents the decision variables of the Hungarian algorithm; Task matching is performed based on the objective function.

[0022] Digital transformation of the Hungarian algorithm, such as Figure 2 As shown, in some embodiments, standardizing the efficiency matrix after task matching includes: Define covering vectors r and c, and matching vector m. Standardize the efficiency matrix. If... Efficiency matrix of order Converting a non-square matrix into a standardized square matrix results in zero cost for the virtual task and a cost for the virtual agent. Then allocation is prohibited, satisfying: in, Define an initial covering vector with zero elements. , , Initial matching vector, For example, its initial value is -1 to indicate no match.

[0023] Based on the standardized efficiency matrix, subtract the minimum value of each row; subtract the minimum value of each column: Subtract the minimum value from each row: Subtract the minimum value from each column: Iterate through all elements ,like ,but That is, assign task i to agent j, and let .

[0024] In some embodiments, standardizing the efficiency matrix after task matching further includes: Determine if the sum of the covering vectors c is equal to n; If the sum of the covering vectors c equals n, then the optimal solution is found, and the set of valid matches is output: {( )}; If the sum of the covering vectors c is not equal to n, then execute: Loop through and check if there are any uncovered 0 elements. but ; If there are no uncovered 0 elements, then calculate the adjustment amount d. ; For all satisfying elements, For all satisfying elements, Reset Cover Vector , Set its elements to 0.

[0025] In some embodiments, updating the drone status, the mission status, and repeating mission matching includes: Task point weight updates at time For each task point i, update its weight as follows: Drone status updates, in real time For each agent j, update its state as follows: in, This indicates the time when agent j was last assigned a task; its meaning is: if a fault is detected, the agent's state is updated to 2; if the current time is greater than or equal to the time when the agent arrived at the current task point, then the task is marked as completed and the drone's state is available; otherwise, the drone's state remains unchanged.

[0026] The efficiency matrix is ​​updated using the improved time efficiency matrix: Update the efficiency matrix elements.

[0027] In some embodiments, updating the drone status, the mission status, and repeating the mission matching further includes: If at time Add a new task k, expand the task point set I = I∪k, and expand the efficiency matrix as follows: The updated objective function is: in, ; The constraints are satisfied: The objective function and constraints are consistent with those of the traditional Hungarian algorithm, and can be solved using the aforementioned digitized Hungarian algorithm.

[0028] Task redistribution is performed using the Hungarian algorithm, and the extended efficiency matrix is ​​redistributed using the digital Hungarian algorithm.

[0029] In some embodiments, it also includes: After the current phase of allocation is completed, update the task allocation results, drone status, and new task status, and perform time increment iterations until the task ends.

[0030] The method in this application considers task urgency and applies timeout penalties when constructing the efficiency matrix. This application digitizes and iterates over time using the traditional Hungarian algorithm to update the drone state, new task state, and efficiency matrix. By applying timeout penalties to the task, it is more suitable for actual rescue missions, particularly in situations where rescue points increase over time and drones may malfunction.

[0031] This application also proposes a multi-agent transportation task matching system, including a processor and a memory. The memory stores a computer program, which, when executed by the processor, implements the steps of the aforementioned multi-agent transportation task matching method.

[0032] It should be noted that, in the embodiments of this application, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0033] The sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0034] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0035] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims. All of these forms are within the protection scope of this application.

Claims

1. A multi-agent transport task matching method, characterized in that, The method comprises: obtaining state information and task demand information of the intelligent agents and initializing an efficiency matrix; performing task matching using a digitized Hungarian algorithm; normalizing the efficiency matrix after task matching; updating the state of the unmanned aerial vehicle and the state of the task and repeating the task matching.

2. The multi-agent transport task matching method of claim 1, wherein, The obtaining state information and task demand information of the intelligent agents and initializing an efficiency matrix comprises: configuring an improved time efficiency matrix; wherein, , i denotes the task number, j denotes the agent, n denotes the maximum value of the agent j in the task number i, denotes the time required for the agent i to complete the task j; denotes the current time, denotes the weight attenuation coefficient, denotes the penalty coefficient for the task i to be completed overtime, denotes the maximum completion time of the task requirement of the task point i, when the time required for the agent j to perform the i task is less than the required completion time of the task i, when the time required for the agent j to perform the i task is greater than the required completion time of the task i, ; Let , initialize the time efficiency matrix as: 。 3. The multi-agent transport task matching method of claim 2, wherein, The performing task matching using a digitized Hungarian algorithm comprises: the constructed objective function is: wherein denotes the decision variable of the Hungarian algorithm; performing task matching based on the objective function.

4. The multi-agent transport task matching method of claim 3, wherein, The normalizing the efficiency matrix after task matching comprises: Define coverage vector r, c, matching vector m, normalize efficiency matrix, if Efficiency matrix of order Convert non-square matrix to normalized square matrix, virtual task cost is 0, virtual agent cost is Then prohibit allocation, meet: wherein Defining an initial coverage vector of 0 elements , , initial matching vector, .

5. based on the normalized efficiency matrix, subtracting the minimum value of each row from the row and the minimum value of each column from the column; Traverse all elements , if , then , i.e., assign task i to agent j and let .

6. The multi-agent transport task matching method of claim 4, wherein, The normalizing the efficiency matrix after task matching further comprises: determining whether the sum of the coverage vector c is equal to n; If the sum of the coverage vectors c equals n, the optimal solution is found and the valid matching set is output: {( }) ; if the sum of the coverage vector c is not equal to n, performing: Loop to check if there is an uncovered 0 element, if then ; If there is no uncovered 0 element, then calculate adjustment d, ; For all elements satisfying ​ For all elements satisfying ​ resetting the cover vector , to have elements equal to 0.

7. The multi-agent transport task matching method of claim 5, wherein, The updating the state of the unmanned aerial vehicle and the state of the task and repeating the task matching comprises: Task point weight update, at time For each task point i, update its weight as: UAV state update, at time For each agent j, update its state as: wherein, represents the time at which agent j was most recently assigned a task; Efficiency matrix update, using improved time efficiency matrix: Update efficiency matrix elements.

8. The multi-agent transport task matching method of claim 5, wherein, The updating the state of the unmanned aerial vehicle and the state of the task and repeating the task matching further comprises: If at time Add task k, extend task point set I = I U k, and extend efficiency matrix as follows: the updated objective function is: wherein ; Satisfy constraints: performing task reassignment using the Hungarian algorithm, and reassigning the extended efficiency matrix using the digitized Hungarian algorithm.

9. The multi-agent transport task matching method of claim 1, wherein, Further comprising: after the end of the current stage assignment, updating the task assignment result, the state of the unmanned aerial vehicle, the new task state, and performing time increment iteration until the end of the task.

10. A multi-agent transport task matching system, comprising: The method comprises a processor and a memory, and the memory stores a computer program, and the computer program is executed by the processor to realize the steps of the multi-intelligent agent transportation task matching method in any one of claims 1 to 8.

Images