WO2022262469A1 - 一种基于博弈论的工业园区物流调度方法及系统 - Google Patents

一种基于博弈论的工业园区物流调度方法及系统 Download PDF

Info

Publication number
WO2022262469A1
WO2022262469A1 PCT/CN2022/091540 CN2022091540W WO2022262469A1 WO 2022262469 A1 WO2022262469 A1 WO 2022262469A1 CN 2022091540 W CN2022091540 W CN 2022091540W WO 2022262469 A1 WO2022262469 A1 WO 2022262469A1
Authority
WO
WIPO (PCT)
Prior art keywords
task
warehouse
freight
freight vehicle
decision
Prior art date
Application number
PCT/CN2022/091540
Other languages
English (en)
French (fr)
Inventor
陈俊华
张珈铜
刘然
黄学达
司凤昊
冉浩宏
Original Assignee
重庆邮电大学工业互联网研究院
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 重庆邮电大学工业互联网研究院 filed Critical 重庆邮电大学工业互联网研究院
Publication of WO2022262469A1 publication Critical patent/WO2022262469A1/zh
Priority to US18/493,598 priority Critical patent/US20240054444A1/en

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G06Q10/047—Optimisation of routes or paths, e.g. travelling salesman problem
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/067—Enterprise or organisation modelling
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/08—Logistics, e.g. warehousing, loading or distribution; Inventory or stock management
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/08—Logistics, e.g. warehousing, loading or distribution; Inventory or stock management
    • G06Q10/083—Shipping
    • G06Q10/0834—Choice of carriers
    • G06Q10/08345—Pricing
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/08—Logistics, e.g. warehousing, loading or distribution; Inventory or stock management
    • G06Q10/083—Shipping
    • G06Q10/0835—Relationships between shipper or supplier and carriers
    • G06Q10/08355—Routing methods
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/08—Logistics, e.g. warehousing, loading or distribution; Inventory or stock management
    • G06Q10/087—Inventory or stock management, e.g. order filling, procurement or balancing against orders

Definitions

  • This application relates to the field of logistics technology and edge computing technology, in particular to a method and system for logistics scheduling in industrial parks based on game theory.
  • Logistics is the process of organically combining functions such as transportation, storage, loading and unloading, packaging, circulation processing, distribution, and information processing according to actual needs during the physical flow of goods from the supply point to the receiving point to realize user needs.
  • the traditional freight dispatching method has problems such as low efficiency and high empty load rate.
  • most of the logistics distribution requests uploaded by the user through the user terminal are adopted.
  • the managers of the logistics department manually dispatch the transport vehicles in the company to respond to the user's logistics distribution request; this scheduling
  • the processing speed of the method is slow, which in turn causes the efficiency of the entire scheduling process to be low.
  • this application proposes a method and system for logistics scheduling in industrial parks based on game theory, using a unified system platform to manage cargo storage information and information in the transportation process, so as to realize user task initiation or acceptance, Functions such as task query and task traceability; according to the Stackelberg game theory, build models for both freight parties to provide optimal pricing and transportation volume decision-making schemes; for problems such as insufficient calculation of freight vehicles in industrial parks, build mobile edge computing nodes to calculate for freight vehicles Optimal decision-making scheme and transportation route planning.
  • this solution significantly improves the transportation efficiency, reduces the empty return rate, and does not affect the real-time performance of tasks, and even improves the real-time performance of task delivery to a certain extent; it has good practical significance and application value .
  • a game theory-based logistics scheduling method for an industrial park comprising:
  • the Stackelberg game model is used to model the warehouse and the freight vehicle, and a logistics scheduling task model is established according to the attributes of the logistics task;
  • the transportation income of the freight vehicle and the storage income of the warehouse corresponding to the logistics task under various decisions are obtained;
  • the optimal decision-making model for each freight vehicle is constructed; with the goal of maximizing storage revenue, an optimal decision-making model for each warehouse is constructed;
  • the optimal decision-making model of the freight vehicle is solved, the solution result is brought into the optimal decision-making model of the warehouse, and the two optimal decision-making models are iteratively solved until the preset threshold is reached, and the respective optimal decision-making models of the freight vehicle and the warehouse are obtained.
  • the optimal decision, and the optimal decision of the freight vehicle and the optimal decision of the warehouse are used for logistics scheduling, which is the decision result of Nash equilibrium.
  • the second aspect of the present application provides a game theory-based industrial park logistics scheduling system
  • the scheduling system includes roadside equipment (RSU) and on-board equipment (OBU), mobile edge computing server (MEC), database and application Platform;
  • RSU roadside equipment
  • OBU on-board equipment
  • MEC mobile edge computing server
  • database and application Platform the scheduling system includes roadside equipment (RSU) and on-board equipment (OBU), mobile edge computing server (MEC), database and application Platform;
  • the roadside equipment provides basic communication support for warehouses and transport vehicles;
  • the mobile edge computing server provides computing support for the transaction process, that is, realizes a game-theory-based industrial Park logistics scheduling method;
  • the database and application platform record task transaction information and broadcast task requests.
  • task participants register an account through the application platform, and each task request will be broadcast to other accounts through the account through the mobile edge computing server; the mobile edge computing server calculates the task model and the Nash equilibrium of each service node, according to the task request
  • a sequence table is obtained in descending order of income, and is continuously updated before the task starts; the sequence table is the priority list of the transaction objects of the task, and the higher the ranking, the higher the priority of participating in the transaction; finally, the entire task process is recorded in In the database, it is convenient for task query and information traceability.
  • This application aims at the cargo scheduling management technology between warehouses and warehouses, warehouses and production lines, and production lines and production lines in industrial parks, and proposes a method and system for logistics scheduling in industrial parks based on game theory to manage cargo storage information and transportation.
  • Information at the same time, it provides the optimal decision-making scheme for both parties of the freight, and selects the transportation route with the lowest cost for the freight vehicle; it does not affect the real-time performance of the task, and even improves the real-time performance of the task delivery to a certain extent; it has good practical significance and application Value; compared with the traditional logistics management system, this application significantly improves the transportation efficiency, reduces the empty return rate, and does not affect the real-time performance of logistics tasks, and even improves the real-time performance of task delivery to a certain extent; it has good practical significance and application value.
  • Fig. 1 is a kind of flow chart of the industrial park logistics dispatching method based on game theory in the embodiment of the present application;
  • Fig. 2 is the logistics scheduling task model figure in the embodiment of the present application.
  • Fig. 3 is a task request flow chart in the embodiment of the present application.
  • Fig. 4 is the task service flowchart in the embodiment of the present application.
  • Fig. 5 is a flow chart of policy analysis after the mobile edge node receives a task in the embodiment of the present application.
  • Fig. 1 is a kind of flow chart of the logistics dispatching method of industrial park based on game theory in the embodiment of the present application, as shown in Fig. 1, described dispatching method comprises:
  • the Stackelberg game model is used to model both parties, and the logistics scheduling task model is established according to the attributes of the logistics task;
  • Cargo resources need to be transported from suppliers to warehouses, sorted and classified, then transported to several factory production lines for processing, and finally the processed finished products are transported to distributors; for each transport task, consider the warehouse as the requester and the warehouse as the requester
  • the pricing relationship between the freight vehicles of the service party In order to obtain greater benefits, freight vehicles need higher pricing, while warehouses need lower pricing in order to obtain greater benefits; this application combines the requester and service
  • the two sides are modeled according to the Stackelberg game; the task initiator, the requester, is the leader of the Stackelberg game, and the task receiver, the server, is the follower of the Stackelberg game.
  • the attribute parameters of the logistics task need to include the quality of the goods, the maximum transportation time limited by the completion of the task, the transportation price, etc., and other unspecified attribute parameters can be specified by those skilled in the art.
  • the logistics task is solved under various decision-making corresponding to the transportation income of the freight vehicle and the storage income of the warehouse;
  • this embodiment considers that the task requester is more inclined to choose a service party with high reputation and has cooperated with itself, so the requester's choice preference for the service party can be defined as the warehouse's subjective preference ⁇ ; the warehouse's Storage income U c can be expressed as:
  • this application considers the timeliness and safety in the task decision-making process.
  • the requester prefers freight vehicles with good delivery records, and also prefers freight vehicles with cooperation experience ;
  • the subjective preference can be composed of multiple weights, subdividing the multi-weight subjective preference, including familiarity weight, time weight and similarity weight.
  • the familiarity weight is defined as Expressed as the relationship frequency between warehouse i and freight vehicle j, the task handover frequency is proportional to the familiarity weight, and the familiarity weight between freight vehicle j and warehouse i can be defined as:
  • f (i, j) represents the number of transactions between warehouse i and freight vehicle j
  • f (i, n) represents the number of transactions between warehouse i and freight vehicle n
  • N is the total number of freight vehicles.
  • f can be expressed as a two-dimensional array and stored in the database. Each transaction can accurately query the value of f[i][j] and bring it into the above formula to obtain the familiar weight of the freight vehicle j corresponding to the current task.
  • the time weight is denoted as It is expressed as the chronological degree of the task relationship between warehouse i and freight vehicle j, and the timestamp of the current time is t. If the task relationship between freight vehicle j and warehouse i is the latest time, then the freight vehicle j will have For larger effects, and vice versa, the time weight between the freight vehicle and the warehouse can be defined as:
  • ⁇ 1 and ⁇ 2 are used to represent the parameters of time influence; Indicates the latest time of the task relationship between freight vehicle j and warehouse i; the transaction time of the recent task is also recorded in the f array, where 0 ⁇ 1 ⁇ 1, ⁇ 2 >1.
  • the similarity weight is denoted as According to the location l j of the warehouse and the location l i of the freight vehicle, the distance between the warehouse and the freight vehicle is determined. A closer distance can not only reduce transportation expenses, but also the requester is more inclined to communicate with other nearby service parties Initiate a transaction.
  • the similarity weight between freight vehicles and warehouses can be defined as:
  • the location of freight vehicle j is It is expressed as the location of the parking lot in the state of no task. If there is a change, the data information needs to be updated.
  • ⁇ 1 + ⁇ 2 + ⁇ 3 1; ⁇ 1 represents the first weight, that is, the familiarity weight, ⁇ 2 represents the second weight, that is, the similarity weight, and ⁇ 3 represents the third weight, that is, the time weight.
  • ⁇ 1 represents the first weight, that is, the familiarity weight
  • ⁇ 2 represents the second weight
  • ⁇ 3 represents the third weight, that is, the time weight.
  • the transportation revenue of freight vehicles can be expressed as:
  • the freight vehicle will serve multiple warehouses at the same time with the goal of increasing its own income;
  • the transportation income of freight vehicles can also be expressed as:
  • ⁇ is a parameter related to the state, which is used to indicate the ratio between the current workload and the maximum workload that can be undertaken, 0 ⁇ 1; W is the depreciation rate of the freight vehicle during transportation; ⁇ represents the current The number of tasks that the freight vehicle needs to perform.
  • this application constructs an optimal decision-making model for each freight vehicle; with the goal of maximizing storage revenue, constructs an optimal decision-making model for each warehouse;
  • the transportation revenue depends on the transportation volume ⁇ i , the pricing ⁇ i and the distance s of the freight vehicle from the starting position to the final position of the task i i .
  • the optimal decision model for each freight vehicle and the optimal decision model for each warehouse are expressed as follows:
  • the utility function and decision variables present the characteristics of convex functions, that is, there is a local optimal solution for a single sub-game, and the process from local optimal to global optimal is the solution process of Nash equilibrium. Equilibrium is the decision result calculated by the optimal decision-making model.
  • ⁇ is a polynomial about the independent variable ⁇ , which is brought into the main game, and the previous calculation is repeated, so that the ⁇ * when the primary derivative of the utility function of the main game is 0 is recorded as:
  • ⁇ * and ⁇ * are the local optimal pricing decision and optimal transportation volume decision, that is, the decision result of the local Nash equilibrium, and also the initial value of the next iterative calculation of the global Nash equilibrium process.
  • the process of solving the Nash equilibrium decision result according to the gradient descent method includes initializing the pricing decision information of the warehouse, and the freight vehicle calculates the transportation volume decision of the freight vehicle through the optimal decision model of the freight vehicle according to the pricing decision information of the warehouse; Based on the transportation volume decision, update the pricing decision of the warehouse using the gradient-assisted search algorithm through the optimal decision model of the warehouse; repeat the iteration until the transportation revenue of the freight vehicle in the current iteration process is the same as that of the freight vehicle in the previous iteration process The revenue change ratio of transportation revenue is less than the preset threshold; output the optimal transportation volume strategy ⁇ * and the optimal pricing strategy ⁇ * at this time.
  • the algorithm running process of gradient descent algorithm for solving Nash equilibrium includes:
  • the pricing strategy ⁇ for the initialization task the number of iterations k ⁇ 1; the preset threshold ⁇ ;
  • the freight vehicle decides the transportation volume decision ⁇ [k] in the k-th iteration process
  • step 6 If Go to step 6, otherwise return to step 2;
  • the game leader strategy that is, the pricing strategy of the warehouse, as the starting point of the gradient descent algorithm, and ⁇ is the policy update band
  • the preset threshold value of income change is used as the basis for the exit condition of the limited number of iterations of the gradient descent algorithm; repeated iterations are performed between steps 2-5 until the preset threshold value condition in step 5 is met; step 2 is to use
  • the convex function characteristics of the utility function make the freight revenue satisfy the local optimum, and a task volume decision can be obtained according to a pricing decision; steps 3-5 represent the specific update process, where k is the number of iterations; step 6 outputs the final Nash Equilibrium, that is, the optimal traffic volume strategy ⁇ * and the optimal pricing strategy ⁇ * .
  • the gradient descent method is used directly.
  • the optimal decision-making model of freight vehicles is solved; otherwise, the multi-task freight path algorithm is used to plan the transportation path, and the execution order of each task is planned under the condition of ensuring that each task is completed within the allowable time, so as to obtain the shortest time before and after participating in the task Path difference.
  • this application proposes a vehicle routing algorithm under multi-task, while ensuring the time allowed for each task Under the condition that the task is completed within , the execution sequence of each task is planned, so as to obtain the total distance difference before and after adding the task, which is the parameter s required for the decision-making process.
  • the planning of the transport route using the multi-task freight route algorithm includes taking the position of the freight vehicle as the origin of the coordinates, the starting and ending positions of each task as the vertices, and generating the distance between the vertices and the origin of the coordinates as a weight
  • the initial matrix diagram based on the dynamic programming algorithm, each vertex set V is expressed in binary, first traverse each vertex, assign the weights of vertices that do not belong to the vertex set, and generate a two-dimensional array of distance weights; Traverse all the vertices in the above vertex set, and update the weights in the array according to the state transition equation; the value of the first row and the last column in the array is the shortest path of the current task i sought, which is the same as the shortest path before receiving the task The difference of the shortest path is the distance from the starting position to the ending position of the freight vehicle performing task i.
  • the running process of the multi-task freight routing algorithm includes:
  • i is from 1 to n;
  • routing arrangements are provided for freight vehicles under multi-task, where the input information is: the task table ⁇ K t , t ⁇ of the freight vehicle, the starting position and target position of each task, and the current position information of the freight vehicle ;
  • Lines 2-4 generate a n*n two-dimensional array arc, and the values in the array represent the distance from the i-th warehouse to the k-th warehouse;
  • lines 5-8 generate an n*2 n-
  • For the two-dimensional array dp of 1 initialize the value of the first column to be equal to the value of the first column of arc; lines 9-19 record the state transition process, and record the weight with the lowest transfer cost into the dp matrix;
  • line 20 outputs s i is the difference of the shortest path before and after participating in the task.
  • the embodiment of the present application also provides a game theory-based industrial park logistics scheduling system, the scheduling system includes roadside equipment, vehicle equipment, mobile edge computing server, database platform and application platform; Provide communication support with freight vehicles; the on-board equipment is used to locate the freight vehicles; the mobile edge computing server provides computing support for the transaction process, that is, implements a logistics scheduling method for industrial parks based on game theory; the database and The application platform records task transaction information and broadcasts task requests.
  • the scheduling system includes roadside equipment, vehicle equipment, mobile edge computing server, database platform and application platform; Provide communication support with freight vehicles; the on-board equipment is used to locate the freight vehicles; the mobile edge computing server provides computing support for the transaction process, that is, implements a logistics scheduling method for industrial parks based on game theory; the database and The application platform records task transaction information and broadcasts task requests.
  • the logistics dispatching system provides task request and task service functions, that is, the warehouse as the requester requests and provides tasks from the freight vehicle as the service party, and the freight vehicle provides the freight task to the corresponding warehouse.
  • the logistics scheduling task model is shown in Figure 2.
  • Cargo resources need to be transported from suppliers to warehouses, sorted and classified, then transported to several factory production lines for processing, and finally the finished products are transported to distributors.
  • Medium-sized freight vehicles are responsible for the transportation of goods between warehouses, and AGVs are responsible for the transportation of goods between the warehouses inside the workshop.
  • AGVs are responsible for the transportation of goods between the warehouses inside the workshop.
  • roadside devices and mobile edge computing nodes are set up in the transportation section.
  • all legal units need to be registered in the application platform, where these legal units include units with transportation task requirements, that is, task request units, such as warehouses, workshop line side warehouses, etc., and service units in freight tasks, namely Task service units such as AGV, large trucks, small trucks, etc.
  • transportation task requirements that is, task request units, such as warehouses, workshop line side warehouses, etc.
  • service units in freight tasks namely Task service units such as AGV, large trucks, small trucks, etc.
  • Fig. 3 is task request flow chart in the embodiment of the present application, as shown in Fig. 3, the process of described task request comprises:
  • the warehouse judges whether there is a demand for goods, that is, whether it needs to purchase goods or ship out goods;
  • Fig. 4 is a flow chart of task service in the embodiment of the present application. As shown in Fig. 4, the process of the task service includes:
  • this application also needs to implement after all legal units are registered on the platform, as follows:
  • Step 1 Register all legal units in the industrial park to the application platform.
  • Step 2 Broadcast each freight task applied by the task request unit to all task service units;
  • the broadcast information should include the ID of the requester, the task completion deadline T, the mass of the transported goods m, the location of the requester L i and the target location L of the goods.
  • Step 3 Freight vehicles parked in the parking lot and not fully loaded will choose to accept the freight task by default after receiving the broadcast information, and send their own information to the adjacent mobile edge computing node; the mobile edge computing node queries the current task table according to each ID information, Calculate the freight task preference degree and Nash equilibrium, and generate the service unit priority queue of the freight task.
  • the self information should include the server ID and the current location L j .
  • Table 1 shows the subjective preference data of the requester.
  • the mobile edge computing node calculates the policy information of each task, the data information is queried first. According to the parameters given by the system and the proportion of each weight, the task’s Subjective preferences, which in turn affect the final outcome of the decision.
  • Step 4 Send the priority queue to the task request unit, and by default conduct a transaction with the freight vehicle at the top of the priority queue, and move backward if the transaction fails or is rejected.
  • Step 5 After the task transaction is completed, the system automatically records the process information and provides a traceability interface.
  • Fig. 5 is a policy analysis process after the mobile edge node receives the task in the embodiment of the present application.
  • 501 Query the task table of the freight vehicle according to the input freight vehicle ID, and add the current freight task to the task table.
  • this application proposes a vehicle routing algorithm under multi-task, which ensures that each task Under the condition of completion within the allowable time, the execution sequence of each task is planned, so as to obtain the total distance difference before and after adding the task, which is the parameter s required for the decision-making process;
  • the specific calculation method includes:
  • the running process of the multi-task freight routing algorithm includes:
  • i is from 1 to n;
  • routing arrangements are provided for freight vehicles under multi-task, where the input information is: the task table ⁇ K t , t ⁇ of the freight vehicle, the starting position and target position of each task, and the current position information of the freight vehicle ;
  • Lines 2-4 generate a n*n two-dimensional array arc, and the values in the array represent the distance from the i-th warehouse to the k-th warehouse;
  • lines 5-8 generate an n*2 n-
  • For the two-dimensional array dp of 1 initialize the value of the first column to be equal to the value of the first column of arc; lines 9-19 record the state transition process, and record the weight with the lowest transfer cost into the dp matrix;
  • line 20 outputs s i is the difference of the shortest path before and after participating in the task.
  • the utility function of the freight car that is, the transportation revenue
  • the utility function is a convex function in the domain of definition, that is, the The function has a maximum value, so that the function equals to the maximum value ⁇ is the solution of the subgame.
  • the pricing strategy ⁇ for the initialization task the number of iterations k ⁇ 1; the preset threshold ⁇ ;
  • the freight vehicle decides the transportation volume decision ⁇ [k] in the k-th iteration process
  • step 6 If Go to step 6, otherwise return to step 2;
  • step 2 is to use the convex function characteristics of the utility function , so that the income satisfies the local optimum, a task volume decision can be obtained according to a pricing decision;
  • step 3-5 represents the specific update process, where k is the number of iterations;
  • step 6 outputs the final Nash equilibrium, that is, the optimal transportation volume Strategy ⁇ * and optimal pricing strategy ⁇ * .
  • the decision information when the preset threshold is reached through a limited number of iterations is the Nash equilibrium; the mobile edge computing node is responsible for recording the utility function value of the current requesting party, and recording all participants in the game according to the value from large to small Service party ID, the queue is the priority of cargo transportation.
  • the freight vehicles participating in the service competition determine their own task list, task execution sequence and service objects.
  • the same warehouse also determines its own request objects, and completes the scheduling tasks planned by the system regularly and quantitatively. Therefore, according to the industrial park logistics management dispatching method and system of the present application, the transportation efficiency is improved, a unified management platform is established, and the safety and traceability of goods are guaranteed to a certain extent, which has good practical significance and application value.

Landscapes

  • Business, Economics & Management (AREA)
  • Engineering & Computer Science (AREA)
  • Human Resources & Organizations (AREA)
  • Economics (AREA)
  • Strategic Management (AREA)
  • Entrepreneurship & Innovation (AREA)
  • General Physics & Mathematics (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Marketing (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Tourism & Hospitality (AREA)
  • Development Economics (AREA)
  • General Business, Economics & Management (AREA)
  • Game Theory and Decision Science (AREA)
  • Educational Administration (AREA)
  • Accounting & Taxation (AREA)
  • Finance (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

一种基于博弈论的工业园区物流调度方法及系统，其包括根据物流任务中仓库和货运车辆的关系，基于博弈论建立出物流调度任务模型；以最大化运输收益为目标构建出每个货运车辆的最优决策模型；以最大化仓储收益为目标构建出每个仓库的最优决策模型；根据梯度下降法，对货运车辆的最优决策模型进行求解，将求解结果输出至仓库的最优决策模型，并迭代两个最优决策模型直至达到预设的阈值，分别得到货运车辆和仓库的最优决策进行物流调度，即为纳什均衡的决策结果。

Description

一种基于博弈论的工业园区物流调度方法及系统
本申请要求于2021年6月18日提交中国专利局、申请号为2021106800250、发明名称为“一种基于博弈论的工业园区物流调度方法及系统”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本申请涉及物流技术与边缘计算技术领域，具体涉及一种基于博弈论的工业园区物流调度方法及系统。
背景技术
物流是物品从供应地向接收地的实体流动过程中，根据实际需要，将运输、储存、装卸搬运、包装、流通加工、配送、信息处理等功能有机结合起来实现用户需求的过程。目前，工业园区中仓库与仓库、仓库与生产线、生产线与生产线之间的货运调度方式单一，传统的货运调度方式存在效率低，空载率较高等问题。现有技术中，大多采用用户通过用户终端上传的多个物流配送请求，物流部门的管理人员接收到物流配送请求后，人工调度公司内的运输车辆响应用户的物流配送请求的方式；这种调度方式处理速度慢，进而造成整个调度过程效率低下。
随着移动边缘计算技术的发展，通过实时分析货运关系能有效提高运输效率。然而当运输货物质量体积远小于货运车的载重时，简单的移动边缘调度方法很有可能造成资源的浪费，多批次的运输任务甚至造成运输过程的拥堵。在工业场景中，多条生产线之间的运输更加复杂，为运输的车辆规划最优的路线方案能有效提高调度效率，如何建立合适的调度模型，结合移动边缘计算的数据分析能力，量化调度效率，对工业园区物流管运输具有重要的意义。
发明内容
基于现有技术存在的问题，本申请提出一种基于博弈论的工业园区物流调度方法及系统，利用统一的系统平台来管理货物仓储信息及运输过程中的信息， 实现用户的任务发起或接受、任务查询和任务追溯等功能；根据Stackelberg博弈理论为货运双方建立模型，提供最优的定价和运输量决策方案；针对工业园区货运车辆计算量不足等问题，构建移动边缘计算节点，为货运车辆计算最优决策方案，以及进行运输路线规划。该方案与传统的物流管理系统相比，运输效率显著提高，降低了空返率，而且不影响任务的实时性，甚至在一定程度上提高任务交付的实时性；具有良好的实际意义和应用价值。
为达到上述目的，本申请提供如下技术方案：
在本申请的第一方面，提供了一种基于博弈论的工业园区物流调度方法，所述方法包括：
根据物流任务中仓库和货运车辆的关系，利用Stackelberg博弈模型对所述仓库和所述货运车辆进行建模，按照物流任务的属性建立出物流调度任务模型；
在所述物流调度任务模型中，根据任务决策依据求解出物流任务在各种决策下对应的货运车辆的运输收益以及仓库的仓储收益；
以最大化运输收益为目标，构建出每个货运车辆的最优决策模型；以最大化仓储收益为目标，构建出每个仓库的最优决策模型；
根据梯度下降法，对货运车辆的最优决策模型进行求解，将求解结果带入仓库的最优决策模型，并迭代求解两个最优决策模型直至达到预设阈值，得到货运车辆和仓库分别的最优决策，并以得到的货运车辆的最优决策和仓库的最优决策进行物流调度，即为纳什均衡的决策结果。
本申请的第二方面，提供了一种基于博弈论的工业园区物流调度系统，所述调度系统包括路侧设备(RSU)和车载设备(OBU)，移动边缘计算服务器(MEC)，数据库和应用平台；所述路侧设备为仓库和运输车辆定位和提供基本的通信支持；所述移动边缘计算服务器为交易过程提供计算支持即实现如本申请第一方面所述的一种基于博弈论的工业园区物流调度方法；所述数据库和应用平台记录任务交易信息和广播任务请求。
具体包括：任务参与者通过应用平台注册账户，每个任务请求会通过该账 户借助移动边缘计算服务器广播到其他账户；移动边缘计算服务器计算该任务模型与每个服务节点的纳什均衡，根据任务请求收益降序得到一个顺序表，并在任务开始前持续更新；所述顺序表即为该任务交易对象的优先级列表，排名靠前的，参与交易的优先级越高；最后将整个任务过程记录到数据库中，方便进行任务查询和信息追溯。
本申请的有益效果：
本申请针对工业园区的仓库与仓库、仓库与生产线、生产线与生产线之间的货物调度管理技术，提出一种基于博弈论的工业园区物流调度方法及系统，来管理货物仓储信息及运输过程中的信息；同时为货运双方提供最优的决策方案，为货运车辆选择成本最低的运输路线；而且不影响任务的实时性，甚至在一定程度上提高任务交付的实时性；具有良好的实际意义和应用价值；本申请与传统的物流管理系统相比，运输效率显著提高，降低了空返率，而且不影响物流任务的实时性，甚至在一定程度上提高任务交付的实时性；具有良好的实际意义和应用价值。
附图说明
为了使本申请的目的、技术方案和有益效果更加清楚，本申请提供如下附图进行说明：
图1是本申请实施例中的一种基于博弈论的工业园区物流调度方法流程图；
图2是本申请实施例中的物流调度任务模型图；
图3是本申请实施例中任务请求流程图；
图4是本申请实施例中任务服务流程图；
图5是本申请实施例中移动边缘节点接收到任务之后的策略分析流程图。
具体实施方式
下面将结合本申请实施例中的附图，对本申请实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例仅仅是本申请一部分实施例，而不是全部的实施例。基于本申请中的实施例，本领域普通技术人员在没有做出创造 性劳动前提下所获得的所有其他实施例，都属于本申请保护的范围。
图1是本申请实施例中的一种基于博弈论的工业园区物流调度方法流程图，如图1所示，所述调度方法包括：
101、根据物流任务中仓库和货运车辆的关系，利用Stackelberg博弈模型对双方进行建模，按照物流任务的属性建立出物流调度任务模型；
货物资源需要从供应商运输到仓库，进行整理分类，再运输到若干工厂生产线上去进行加工，最后将加工后的成品运输到经销商；针对每一个运输任务，考虑到作为请求方的仓库和作为服务方的货运车辆之间的定价关系，货运车辆为了获得更大的利益，则需要更高的定价，而仓库为了获得更大的利益，则需要更低的定价；本申请将请求方和服务方作为博弈双方，根据Stackelberg博弈对双方进行建模；任务发起者即请求方作为Stackelberg博弈的领导者，任务接收者即服务方作为Stackelberg博弈的跟随者。
其中，物流任务的属性参数需要包括货物质量，完成任务限制的最大运输时间，运输价格等，其他未被指明的属性参数可由本领域技术人员进行指定。
102、在所述物流调度任务模型中，根据任务决策依据求解出物流任务在各种决策下对应货运车辆的运输收益以及仓库的仓储收益；
作为Stackelberg博弈的领导者，仓库首先作出对本批次货物定价的策略{λ＝[λ i]i∈N：λ min＜λ i}，这个定价策略将作为服务费用，其中λ min为最低定价，从公式看出，仓库对任务的定价必须满足大于最低定价λ min，λ i为仓库对任务i的定价，货运车辆按照仓库对任务i的定价执行任务i，λ i也为货运车辆执行任务i的定价，N为总的任务数量；假设本次货运货物的总量为m，运送每单位的货物对仓库产生的收益记为R。
在本发明实施例中，本实施例认为任务请求方更偏向选择信誉高、与自身有过合作的服务方，所以可以将请求方对服务方的选择偏好定义为仓库的主观偏好ω；仓库的仓储收益U c可以表达为：
U c＝(Rμ i-λ i)ω；
在上述实施例中，本申请考虑到任务决策过程中的时效性和安全性，对于工业园区而言，请求方更偏向于有良好交付记录的货运车辆，也偏向于有过合作经历的货运车辆；那么主观偏好就可以由多权重构成，对多权重主观偏好进行细分，包括熟悉权重、时间权重和相似权重。
所述熟悉权重定义为
Figure PCTCN2022091540-appb-000001
表现为仓库i和货运车辆j之间的关系频率，任务交接频率正比于熟悉权重，货运车辆j与仓库i之间的熟悉权重可以定义为：
Figure PCTCN2022091540-appb-000002
其中，f (i,j)表示当前仓库i与货运车辆j的交易次数；f (i,n)表示当前仓库i与货运车辆n的交易次数；N是总的货运车辆的数量。f可以表现为一个二维数组存储在数据库中，每次交易可以通过精确查询f[i][j]的值，带入上式求得该货运车辆j对应当前任务的熟悉权重。
所述时间权重记为
Figure PCTCN2022091540-appb-000003
表现为仓库i和货运车辆j发生任务关系的时间先后程度，当前时间的时间戳记为t，，如果货运车辆j和仓库i的任务关系是最近的时间，那么该货运车辆j会对仓库i有比较大的影响，反之亦然，货运车辆与仓库之间的时间权重可以定义为：
Figure PCTCN2022091540-appb-000004
其中，α 1和α 2用来表示时间影响的参数；
Figure PCTCN2022091540-appb-000005
表示货运车辆j和仓库i的发生任务关系的最近时间；较近任务的交易时间同样记录在f数组中，其中0＜α 1＜1，α 2＞1。
所述相似权重记为
Figure PCTCN2022091540-appb-000006
根据仓库所处的位置l j和货运车辆所处的位置l i，确定出仓库和货运车辆之间的距离，距离更近不但能减少运输开支，请求方也更倾向于与附近的其他服务方发起交易。货运车辆与仓库之间的相似权重可以定义为：
Figure PCTCN2022091540-appb-000007
其中，货运车辆j所处的位置
Figure PCTCN2022091540-appb-000008
表现为没有任务状态下所在的停车厂位置，如果发生变动需要更新数据信息。
综合考虑到熟悉权重、时间权重和相似权重，可以得到当前仓库对货运车辆的主观偏好，
Figure PCTCN2022091540-appb-000009
其中γ 1+γ 2+γ 3＝1；γ 1表示第一权重即熟悉权重，γ 2表示第二权重即相似权重，γ 3表示第三权重即时间权重，这些权重的取值可以根据现有技术进行赋予。
作为Stackelberg博弈的跟随者，货运车辆制定运输量策略，确定其效用函数；对每辆货运车辆的效用函数进行分析，定义其策略为运输量μ i，{μ＝[μ i]i∈N：0＜μ i＜M}，μ i表示货运车辆执行任务i的运输量；货运过程中单位里程的花费包括燃油、损耗，记为c，从起始点到终点的路程记为s i，M为货运车辆的最大载货量。
所以，货运车辆的运输收益U s可以分为以下两种情况：
如果当前货运车辆的运输量M小于仓库的运输量m，则需要多辆货运车辆协同完成任务；
货运车辆的运输收益可以表示为：
U s＝μ iλ i-s ic-β(μ iW) 2
如果当前货运车辆的运输量M大于仓库的运输量m，该货运车辆以提高自身收益为目标，会同时为多个仓库服务；
所以，货运车辆的运输收益还可以表示为：
Figure PCTCN2022091540-appb-000010
其中，β是一个与状态相关的参数，用于指示当前工作负载与可以承担的最大工作负载之间的比率，0＜β≤1；W是货运车辆在运输过程中的折旧率；Φ表示当前货运车辆需要执行的任务数。
103、以最大化运输收益为目标，本申请构建出每个货运车辆的最优决策模 型；以最大化仓储收益为目标，构建出每个仓库的最优决策模型；
为了让仓储收益最大化，由于针对某特定的仓库，R属于一个定值，因此需要运输量μ i最大，且定价λ i最小。
为了让运输收益最大化，由于针对某特定的货运车辆，W属于一个定值，因此运输收益取决于运输量μ i、定价λ i以及货运车辆执行任务i从起始位置到终止位置的路程s i。
基于上述分析，每个货运车辆的最优决策模型和每个仓库的最优决策模型表示如下：
Figure PCTCN2022091540-appb-000011
Figure PCTCN2022091540-appb-000012
Figure PCTCN2022091540-appb-000013
表示货运车辆执行任务i的最优运输量决策；
Figure PCTCN2022091540-appb-000014
表示货运车辆执行任务i的最优定价决策。
104、根据梯度下降法，对货运车辆的最优决策模型进行求解，将求解结果输出至仓库的最优决策模型，并迭代两个最优决策模型直至达到预设的阈值，得到货运车辆和仓库分别的最优决策进行物流调度，即为纳什均衡的决策结果。
通过物流调度任务模型可以发现，效用函数和决策变量呈现出凸函数的特点，即单个子博弈存在局部最优解，从局部最优到全局最优的过程为纳什均衡的求解过程，所述纳什均衡即为最优决策模型所计算出的决策结果。
为了证明纳什均衡的存在性和唯一性，针对双方的效用函数即收益函数关于变量λ，μ做出以下分析，首先求解子博弈的局部最优，根据凸函数的特点，求出使得所述效用函数一次导数为0时的μ *，即为：
Figure PCTCN2022091540-appb-000015
Figure PCTCN2022091540-appb-000016
由上式可以发现，μ是关于自变量λ的多项式，将其带入到主博弈中，重复之前的运算，使得主博弈效用函数一次导数为0时的λ *，记为：
Figure PCTCN2022091540-appb-000017
Figure PCTCN2022091540-appb-000018
其中，μ *，λ *为局部局部最优定价决策和最优运输量决策，也就是局部纳什均衡的决策结果，也是下一步迭代计算全局纳什均衡过程的初值。
具体的，根据梯度下降法求解出纳什均衡的决策结果的过程包括初始化仓库的定价决策信息，货运车辆根据仓库的定价决策信息通过货运车辆的最优决策模型，计算出货运车辆的运输量决策；基于所述运输量决策，通过仓库的最优决策模型对仓库的定价决策使用梯度辅助搜索算法进行更新；重复迭代，直至当前迭代过程中货运车辆的运输收益与前一轮迭代过程的货运车辆的运输收益的收益变化比例小于预设阈值；输出此时的最优运输量策略μ *和最优定价策略λ *。
在一些优选实施例中，梯度下降算法求解纳什均衡的算法运行过程包括：
1、初始化任务的定价策略λ，迭代次数k←1；预设阈值ε；
2、货运车辆根据仓库请求方的定价策略，决策出第k轮迭代过程中的运输量决策μ [k]；
3、对仓库的定价策略使用梯度辅助搜索算法，即采用
Figure PCTCN2022091540-appb-000019
更新；其中v表示仓库策略更新的步数，定价策略信息会同步发送到所有的货运车辆节点。
4、令迭代次数k自加1，表示为k←k+1；
5、若
Figure PCTCN2022091540-appb-000020
进入步骤6，否则返回步骤2；
6、输出最优运输量μ *[k]和最优定价λ *[k]。
上述实施例中，进行初始化之前，还需要输入双方的效用函数即双方的收益函数U c和U s，初始化博弈领导者策略即仓库的定价策略，作为梯度下降算法的起点，ε为策略更新带来收益变化的预设阈值，其作为梯度下降算法的有限次迭代退出条件的依据；在步骤2-5之间进行反复迭代，直到满足步骤5中的预设阈值条件退出；其中步骤2是利用效用函数的凸函数特点，使货运收益满足局部最优，可以根据一个定价决策，求得一个任务量决策；步骤3-5表示具体的更新过程，其中k为迭代次数；步骤6输出最终的纳什均衡，即最优运输量策略μ *和最优定价策略λ *。
在一些实施例中，考虑到单个货运车辆可以服务多个货运任务也可以服务单个货运任务，因此本实施例中还可以判断当前的物流任务是否大于1，若大于1则直接梯度下降法，对货运车辆的最优决策模型进行求解；否则采用多任务货运路径算法规划运输路径，在保证每个任务允许时间内完成的条件下，规划了每个任务的执行顺序，从而得到参与该任务前后最短路径的差值。
针对多辆运输车竞争同一任务的问题，需要对每个货运车求解当前任务的纳什均衡解，对每个均衡条件下的收益进行排序，即可得到任务对象的优先级。默认选择优先级最高的货运车来执行本次任务。
考虑到多任务的起始点和目标点可能重叠或者路程交叉往复，为提高运输效率，为解决所述车辆路径问题，本申请提出一种多任务下的车辆路径算法，在保证每个任务允许时间内完成的条件下，规划了每个任务的执行顺序，从而得到添加该任务前后的总路程差值，即为决策过程所需要的参数s。
在一些实施例中，所述采用多任务货运路径算法规划运输路径包括以货运车辆位置为坐标原点，每个任务的起始、终止位置为顶点，将顶点与坐标原点之间的距离作为权重生成初始矩阵图；基于动态规划算法，用二进制表示各顶点集合V，首先遍历每个顶点，对不属于该顶点集合的顶点的权值进行赋值， 生成关于路程权值的二维数组；再对所述顶点集合中所有的顶点进行遍历，根据状态转移方程，更新数组中的权值；该数组中第一行最后一列的值即为所求的当前任务i的最短路径，与接收该任务之前的最短路径的差值即为货运车辆执行任务i从起始位置到终止位置的路程。
在一些优选实施例中，多任务货运路径算法运行过程包括：
1、输入：仓库t的位置L(x t，y t)以及该货运车辆的位置L，该货运车辆任务表K{K t，t}；
2、以该货运车辆位置为坐标原点，以任务的起始位置和终点位置为顶点，将仓库位置坐标L(x t，y t)之间的距离作为权重生成整张图；
3、初始化图的代价矩阵arc[n][n]，该代价矩阵为一个n×n大小的二维数组；
4、计算出仓库t与另一个仓库k之间的距离，并存储到所述代价矩阵中
Figure PCTCN2022091540-appb-000021
5、初始化所述代价矩阵中的第一列，即货运车辆到该仓库t之间的距离；
6、对于代价矩阵中任意一列，令i从1到n；
7、令dp[i][0]＝arc[t][0]；
8、执行步骤6-7；
9、对于代价矩阵中任意一行，令j从1到2 n-1-1；
10、对于代价矩阵中的任意一列，i从1到n；
11、当仓库i不属于货运车辆j的目标仓库集合；
12、对任意的仓库k从1到n；
13、当仓库k属于货运车辆j的目标仓库集合；
14、令dp[i][j]＝min{arc[t][k]+dp[i][j-2 (k-1)]，dp[i][j]}；
15、完成步骤13-14的循环；
16、完成步骤12-15的循环；
17、完成步骤11-16的循环；
18、完成步骤10-17的循环；
19、完成步骤9-18的循环；
20、输出s i＝d[0][2 n-1-1]–s i-1。
本实施例中，为多任务下货运车辆提供了路径安排，其中输入信息为：该货运车辆的任务表{K t，t}，每个任务的起始位置、目标位置以及货运车辆当前位置信息；第2-4行生成了一个n*n的二维数组arc，数组中的值代表了第i个仓库到第k个仓库的路程长度；第5-8行生成了一个n*2 n-1的二维数组dp，初始化第一列的值等于arc第一列的值；第9-19行记录了状态转移过程，将转移代价最低的权值记录到dp矩阵中；第20行输出s i，即参与该任务前后最短路径的差值。
本申请实施例中还提供了一种基于博弈论的工业园区物流调度系统，所述调度系统包括路侧设备、车载设备、移动边缘计算服务器、数据库平台以及应用平台；所述路侧设备对仓库和货运车辆提供通信支持；所述车载设备用于对所述货运车辆定位；所述移动边缘计算服务器为交易过程提供计算支持即实现一种基于博弈论的工业园区物流调度方法；所述数据库和应用平台记录任务交易信息和广播任务请求。
在本申请实施例中，所述物流调度系统提供任务请求和任务服务功能，即作为请求方的仓库向作为服务方的货运车辆请求提供任务，所述货运车辆向对应的仓库提供货运任务。
物流调度任务模型如图2所示。货物资源需要从供应商运输到仓库，进行整理分类，再运输到若干工厂生产线上去进行加工，最后将成品运输到经销商。中型货运车辆负责仓库与仓库之间的货物运输，AGV负责车间内部线边仓之间的货物运输。为了满足运输车辆之间的通信和数据处理，在运输路段设置若干路侧设备和移动边缘计算节点。
在本申请实施例中，需要在应用平台中注册所有合法单元，其中这些合法单元包括有运输任务需求的单元即任务请求单元，例如仓库，车间线边仓等以及在货运任务中的服务单元即任务服务单元例如AGV，大货车，小货车等。
图3是本申请实施例中任务请求流程图，如图3所示，所述任务请求的过 程包括：
201、仓库判断自身是否存在货物需求，即是否需要采购货物或者运出货物；
202、若存在货物需求，则向任务请求单元提交任务申请；
203、经过移动边缘计算服务器分析物流调度任务模型，并获得任务申请的在任务队列中的优先级；
204、根据移动边缘计算服务器的分析结果，确认交易。
图4是本申请实施例中任务服务流程图，如图4所示，所述任务服务的过程包括：
301、判断任务队列是否空闲或者非满载；
302、若任务队列有空闲，则接收其他任务请求单元提交的任务申请的广播；
303、经过移动边缘计算服务器分析物流调度任务模型，得到该任务申请在任务队列中的位置；
304、若处于任务队列中的最靠前的位置，则根据移动边缘计算服务器的分析结果，确认交易。
在一些实施例中，本申请还需要将所有合法单元注册到平台后进行实施，如下：
步骤1：对工业园区所有合法单元注册到应用平台中。
步骤2：将任务请求单元申请的每个货运任务广播到所有任务服务单元；
其中，广播信息应包括请求方ID，任务完成截止时间T，运输货物质量m，请求方位置L i以及货物的目标位置L。
步骤3：停靠在停车场和非满载的货运车辆接收到广播信息后默认选择接受货运任务，并将自身信息发送到邻近的移动边缘计算节点；移动边缘计算节点根据各个ID信息查询当前任务表，计算货运任务偏好程度以及纳什均衡，生成该货运任务的服务单元优先级队列。
其中，所述自身信息应包括该服务方ID，当前所处位置L j。
表1为请求方主观偏好数据，在移动边缘计算节点计算每次任务的策略信 息时，先查询所述的数据信息，根据系统给定的参数和各项权重的占比，得出该任务的主观偏好，进而影响决策的最终结果。
表1 请求方主观偏好数据
  请求方 服务方
当前位置 L(x i,y i) L(x j,y j)
交易次数 f(i,j) f(i,j)
上次交易时间 t(i,j) t(i,j)
步骤4：将所述优先级队列发送到任务请求单元，默认与处于优先级队列最前的货运车辆进行交易，如交易失败或者拒绝，则向后顺移。
步骤5：任务交易完成，系统自动记录过程信息，并提供追溯功能接口。
图5是本申请实施例中移动边缘节点接收到任务之后的策略分析流程。对于任务数据处理，及纳什均衡求解，主要包括以下步骤：
501、根据输入的货运车辆ID查询该货运车辆的任务表，将当前货运任务添加到任务表中。
502、考虑到多任务的起始点和目标点可能重叠或者路程交叉往复，为提高运输效率，为解决所述车辆路径问题，本申请提出一种多任务下的车辆路径算法，在保证每个任务允许时间内完成的条件下，规划了每个任务的执行顺序，从而得到添加该任务前后的总路程差值，即为决策过程所需要的参数s；具体计算方法包括：
在一些优选实施例中，多任务货运路径算法运行过程包括：
1、输入：仓库t的位置L(x t，y t)以及该货运车辆的位置L，该货运车辆任务表K{K t，t}，
2、以该货运车辆位置为坐标原点，以任务的起始位置和终点位置为顶点，将仓库位置坐标L(x t，y t)之间的距离作为权重生成整张图；
3、初始化图的代价矩阵arc[n][n]，该代价矩阵为一个n×n大小的二维数组；
4、计算出仓库t与另一个仓库k之间的距离，并存储到所述代价矩阵中
Figure PCTCN2022091540-appb-000022
5、初始化所述代价矩阵中的第一列，即货运车辆到该仓库t之间的距离；
6、对于代价矩阵中任意一列，令i从1到n；
7、令dp[i][0]＝arc[t][0]；
8、执行步骤6-7；
9、对于代价矩阵中任意一行，令j从1到2 n-1-1；
10、对于代价矩阵中的任意一列，i从1到n；
11、当仓库i不属于货运车辆j的目标仓库集合；
12、对任意的仓库k从1到n；
13、当仓库k属于货运车辆j的目标仓库集合；
14、令dp[i][j]＝min{arc[t][k]+dp[i][j-2 (k-1)]，dp[i][j]}；
15、完成步骤13-14的循环；
16、完成步骤12-15的循环；
17、完成步骤11-16的循环；
18、完成步骤10-17的循环；
19、完成步骤9-18的循环；
20、输出s i＝d[0][2 n-1-1]–s i-1。
本实施例中，为多任务下货运车辆提供了路径安排，其中输入信息为：该货运车辆的任务表{K t，t}，每个任务的起始位置、目标位置以及货运车辆当前位置信息；第2-4行生成了一个n*n的二维数组arc，数组中的值代表了第i个仓库到第k个仓库的路程长度；第5-8行生成了一个n*2 n-1的二维数组dp，初始化第一列的值等于arc第一列的值；第9-19行记录了状态转移过程，将转移代价最低的权值记录到dp矩阵中；第20行输出s i，即参与该任务前后最短路径的差值。
503、根据步骤502得到的路程差值s，可以精确定义货运车的效用函数即运输收益，是关于自变量μ的二次多项式，显然所述效用函数在定义域上呈现为凸函数，即该函数存在极大值，使得该函数等于极大值的μ即为子博弈的解。
504、由跟随者决策反推领导者决策，依次迭代；迭代更新是指策略使得领导者的效用函数最大化的过程，根据效用函数的一次导数更新策略；由所述效用函数可以得到一个局部最优解；接收到货运车的运输量策略后，仓库做出新一轮的定价策略，随后货运车再根据该策略制定新的策略；反复迭代直到达到纳什均衡。具体计算方法包括：
1、初始化任务的定价策略λ，迭代次数k←1；预设阈值ε；
2、货运车辆根据仓库请求方的定价策略，决策出第k轮迭代过程中的运输量决策μ [k]；
3、对仓库的定价策略使用梯度辅助搜索算法，即采用
Figure PCTCN2022091540-appb-000023
更新；其中v表示仓库策略更新的步数，定价策略信息会同步发送到所有的货运车辆节点。
5、令迭代次数自加1，表示为k←k+1；
5、若
Figure PCTCN2022091540-appb-000024
进入步骤6，否则返回步骤2；
6、输出最优运输量μ *[k]和最优定价λ *[k]。
上述实施例中，进行初始化之前，还需要输入是双方的效用函数即双方的收益函数U c和U s，初始化博弈领导者策略即仓库的定价策略，作为梯度下降算法的起点，ε为策略更新带来收益变化的阈值，作为有限次迭代退出条件的依据；在步骤2-5之间进行反复迭代，直到满足步骤5中的预设阈值条件退出；其中步骤2是利用效用函数的凸函数特点，使收益满足局部最优，可以根据一个定价决策，求得一个任务量决策；步骤3-5表示具体的更新过程，其中k为迭代次数；步骤6输出最终的纳什均衡，即最优运输量策略μ *和最优定价策略λ *。
505、通过有限次数的迭代达到预设阈值时的决策信息，即为纳什均衡；移动边缘计算节点负责记录当前请求方的效用函数值，按照该值从大到小排列记录所有参与到博弈中的服务方ID，该队列即为货物运输的优先级。
经过移动边缘节点的一轮计算，参与到服务竞争的货运车辆确定了自身的任务表，任务执行顺序和服务对象，同样的仓库也确定了自身的请求对象，定时定量完成系统规划的调度任务。由此，根据本申请的工业园区物流管理调度方法及系统，提高了运输效率，建立了统一的管理平台，一定程度上保障货物的安全可追溯，具有良好的实际意义和应用价值。
尽管已经示出和描述了本申请的实施例，对于本领域的普通技术人员而言，可以理解在不脱离本申请的原理和精神的情况下可以对这些实施例进行多种变化、修改、替换和变型，本申请的范围由所附权利要求及其等同物限定。

Claims (9)

  1. 一种基于博弈论的工业园区物流调度方法，其特征在于，包括：
    根据物流任务中仓库和货运车辆的关系，利用Stackelberg博弈模型对所述仓库和所述货运车辆双方进行建模，按照物流任务的属性建立出物流调度任务模型；
    在所述物流调度任务模型中，根据任务决策依据求解出物流任务在各种决策下对应的货运车辆的运输收益以及仓库的仓储收益；
    以最大化运输收益为目标，构建出每个货运车辆的最优决策模型；以最大化仓储收益为目标，构建出每个仓库的最优决策模型；
    根据梯度下降法，对货运车辆的最优决策模型进行求解，将求解结果带入仓库的最优决策模型中，并迭代求解两个最优决策模型直至达到预设阈值，分别得到货运车辆和仓库的最优决策，并以得到的货运车辆的最优决策和仓库的最优决策进行物流调度，即为纳什均衡的决策结果。
  2. 根据权利要求1所述的一种基于博弈论的工业园区物流调度方法，其特征在于，所述仓库的仓储收益的计算方式表示为：
    U c＝(Rμ i-λ i)ω；
    其中，U c表示仓库的仓储收益，R表示运送每单位的货物对仓库产生的收益；μ i表示货运车辆执行任务i的运输量；λ i表示货运车辆执行任务i的定价；ω表示仓库对货运车辆的主观偏好。
  3. 根据权利要求2所述的一种基于博弈论的工业园区物流调度方法，其特征在于，所述仓库对货运车辆的主观偏好通过货运车辆与仓库之间的熟悉权重，货运车辆与仓库之间的时间权重，以及货运车辆与仓库之间的相似权重进行加权求和得到。
  4. 根据权利要求1所述的一种基于博弈论的工业园区物流调度方法，其特征在于，所述货运车辆的运输收益的计算方式包括：
    若当前货运车辆的运输量M小于仓库的运输量m，则需要多辆货运车协同完成任务，货运车辆的运输收益表示为：
    U s＝μ iλ i-s ic-β(μ iW) 2
    若当前货运车辆的运输量M大于仓库的运输量m，该货运车辆同时为多个仓库服务，货运车辆的运输收益表示为：
    Figure PCTCN2022091540-appb-100001
    其中，U s表示货运车辆的运输收益；μ i表示货运车辆执行任务i的运输量；λ i表示货运车辆执行任务i的定价；Φ表示当前货运车辆需要执行的物流任务数；c表示货运车辆在货运过程中单位里程的花费；s i表示货运车辆执行任务i从起始点到终点的路程；β是一个与状态相关的参数，用于指示当前工作负载与可以承担的最大工作负载之间的比率；W是货运车辆在运输过程中的折旧率。
  5. 根据权利要求1所述的一种基于博弈论的工业园区物流调度方法，其特征在于，在采用梯度下降法求解出纳什均衡的决策结果之前还包括判断当前的物流任务是否大于1，若大于1则采用多任务货运路径算法规划运输路径；在保证每个任务允许时间内完成的条件下，规划每个任务的执行顺序，按照任务的执行顺序得到执行任务的最短路径；根据梯度下降法，对货运车辆的最优决策模型进行求解，得到每个任务的决策信息。
  6. 根据权利要求5所述的一种基于博弈论的工业园区物流调度方法，其特征在于，所述采用多任务货运路径算法规划运输路径包括以货运车辆位置为坐标原点，每个任务的起始、终止位置为顶点，将顶点与坐标原点之间的距离作为权重生成初始矩阵图；基于动态规划算法，用二进制表示各顶点集合V，首先遍历每个顶点，对不属于该顶点集合的顶点的权值进行赋值，生成关于路程权值的二维数组；再对所述顶点集合中所有的顶点进行遍历，根据状态转移方程，更新数组中的权值；该数组中第一行最后一列的值即为所求的当前任务i的最短路径，与接收该任务之前的最短路径的差值即为货运车辆执行任务i从起始 位置到终止位置的路程。
  7. 根据权利要求1所述的一种基于博弈论的工业园区物流调度方法，其特征在于，根据梯度下降法求解出纳什均衡的决策结果的过程包括初始化仓库的定价决策信息，货运车辆根据仓库的定价决策信息通过货运车辆的最优决策模型，计算出货运车辆的运输量决策；基于所述运输量决策，通过仓库的最优决策模型对仓库的定价决策使用梯度辅助搜索算法进行更新；重复迭代，直至当前迭代过程中货运车辆的运输收益与前一轮迭代过程的运输收益的收益变化比例小于预设阈值；输出此时的最优运输量策略μ *和最优定价策略λ *。
  8. 根据权利要求1所述的一种基于博弈论的工业园区物流调度方法，其特征在于，所述方法还包括当多个货运车辆竞争同一物流任务时，对每个货运车辆求解出当前任务的纳什均衡解，对每个均衡条件下的运输收益进行排序，得到任务对象的优先级，并选择出优先级最高的货运车辆执行本次任务。
  9. 一种基于博弈论的工业园区物流调度系统，其特征在于，所述调度系统包括路侧设备、车载设备、移动边缘计算服务器、数据库平台以及应用平台；所述路侧设备对仓库和货运车辆提供通信支持；所述车载设备用于对所述货运车辆定位；所述移动边缘计算服务器为交易过程提供计算支持即实现如权利要求1～8任一所述的一种基于博弈论的工业园区物流调度方法；所述数据库和应用平台记录任务交易信息和广播任务请求。
PCT/CN2022/091540 2021-06-18 2022-05-07 一种基于博弈论的工业园区物流调度方法及系统 WO2022262469A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US18/493,598 US20240054444A1 (en) 2021-06-18 2023-10-24 Logistics scheduling method and system for industrial park based on game theory

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202110680025.0A CN113393040B (zh) 2021-06-18 2021-06-18 一种基于博弈论的工业园区物流调度方法及系统
CN202110680025.0 2021-06-18

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US18/493,598 Continuation US20240054444A1 (en) 2021-06-18 2023-10-24 Logistics scheduling method and system for industrial park based on game theory

Publications (1)

Publication Number Publication Date
WO2022262469A1 true WO2022262469A1 (zh) 2022-12-22

Family

ID=77623088

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/091540 WO2022262469A1 (zh) 2021-06-18 2022-05-07 一种基于博弈论的工业园区物流调度方法及系统

Country Status (3)

Country Link
US (1) US20240054444A1 (zh)
CN (1) CN113393040B (zh)
WO (1) WO2022262469A1 (zh)

Cited By (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116108550A (zh) * 2023-04-13 2023-05-12 湖北工业大学 一种基于bim的多式联运的动态优化方法及系统
CN116187895A (zh) * 2023-04-24 2023-05-30 北京中超伟业信息安全技术股份有限公司 一种智能仓储货流规划方法、系统及电子设备
CN116342197A (zh) * 2023-01-05 2023-06-27 上海朗晖慧科技术有限公司 一种结合地理位置信息的实时价格方案生成方法
CN116823124A (zh) * 2023-08-30 2023-09-29 四川普鑫物流自动化设备工程有限公司 一种集中式多层货架四向车调度系统及方法
CN116873431A (zh) * 2023-07-07 2023-10-13 湘南学院 一种基于岩板智能仓库的多重载agv储运方法
CN117196262A (zh) * 2023-11-06 2023-12-08 中船凌久高科(武汉)有限公司 一种基于状态编码优化的测试场车辆与场景匹配调度方法
CN117455341A (zh) * 2023-12-25 2024-01-26 华东交通大学 一种智能化车货匹配灵活调优方法
CN117808596A (zh) * 2024-01-26 2024-04-02 国金证券股份有限公司 证券期货行业基于长短期记忆模块的大语言模型多Agent系统
CN117933882A (zh) * 2024-02-01 2024-04-26 山东曼索信息科技有限公司 一种基于物联网的智能物流系统
CN118313736A (zh) * 2024-06-07 2024-07-09 张家港大裕橡胶制品有限公司 用于橡胶手套生产的复合产线加工管理方法及系统

Families Citing this family (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113393040B (zh) * 2021-06-18 2023-04-07 重庆邮电大学工业互联网研究院 一种基于博弈论的工业园区物流调度方法及系统
CN114331220B (zh) * 2022-03-01 2022-05-13 北京邮电大学 基于订单动态优先级的乘用车运输车辆调度方法及装置
CN114663012B (zh) * 2022-03-10 2024-06-25 重庆邮电大学 一种基于区块链的生产物流管理方法
CN115311846B (zh) * 2022-06-24 2023-08-11 华东师范大学 一种结合货车任务状态的厂区道路拥堵预测方法及预测系统
CN116107276A (zh) * 2022-12-30 2023-05-12 福州大学 基于分布式微分博弈的物流仓储最优协调方法
CN116542458B (zh) * 2023-04-28 2024-02-23 北京大数据先进技术研究院 一种载具分配方法、系统和电子设备
CN117973989B (zh) * 2024-04-02 2024-08-13 安徽轩创智能科技有限公司 一种基于5g技术的智造产业园管理系统
CN118689186A (zh) * 2024-08-23 2024-09-24 珠海格力电器股份有限公司 基于知识图谱的agv辅助调度方法及装置

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20070156883A1 (en) * 2005-12-30 2007-07-05 Visual Exchange Network, Llc System and method for group to group entertainment
CN107730091A (zh) * 2017-09-26 2018-02-23 沈阳工业大学 一种单周期二级物流服务供应链的协调运作方法
CN108171459A (zh) * 2017-12-29 2018-06-15 长春师范大学 基于博弈论的智能仓储优化方法
CN110378577A (zh) * 2019-07-02 2019-10-25 武汉理工大学 一种基于Supply-Hub的供应商管理库存运作方法
US20190347371A1 (en) * 2018-05-09 2019-11-14 Volvo Car Corporation Method and system for orchestrating multi-party services using semi-cooperative nash equilibrium based on artificial intelligence, neural network models,reinforcement learning and finite-state automata
CN111199375A (zh) * 2020-01-13 2020-05-26 江苏佳利达国际物流股份有限公司 一种智能化物流运输系统
US20200167722A1 (en) * 2018-11-27 2020-05-28 Zume Inc. Delivery of food items by aerial or ground drones to and from delivery vehicles
CN113393040A (zh) * 2021-06-18 2021-09-14 重庆邮电大学工业互联网研究院 一种基于博弈论的工业园区物流调度方法及系统

Family Cites Families (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8014809B2 (en) * 2006-12-11 2011-09-06 New Jersey Institute Of Technology Method and system for decentralized power control of a multi-antenna access point using game theory
CN104036403A (zh) * 2013-03-06 2014-09-10 江苏润和软件股份有限公司 一种基于重复博弈论的供应链网络分级评估系统
CN104506288A (zh) * 2015-01-23 2015-04-08 重庆邮电大学 基于斯塔克伯格博弈的概率网络编码重传方法
CN106960246A (zh) * 2017-03-17 2017-07-18 重庆邮电大学 一种基于演化博弈的车辆诱导方法
CN106991555B (zh) * 2017-03-29 2021-01-01 北京理工大学 基于激励机制的城市车辆众包包裹传递方法
CN108470226B (zh) * 2018-03-27 2024-01-12 山东高速信联支付有限公司 一种物流系统收益最大化方法
CN109495906B (zh) * 2018-11-09 2022-04-26 重庆邮电大学 基于无人机-地面站链路预测的无人机网关选择方法
CN109978458B (zh) * 2019-03-28 2023-04-21 广西大学 物流运输业务快速协同响应的决策信息可视化表达方法
CN112488740A (zh) * 2019-09-12 2021-03-12 清华大学 目标均衡博弈的处理方法和装置
CN110751293B (zh) * 2019-09-29 2022-04-15 浙江财经大学 一种基于博弈论的云制造多任务调度优化方法
CN110659839A (zh) * 2019-09-29 2020-01-07 杭州货天下物流科技有限公司 一种物流智能配载调度方法
CN110826958A (zh) * 2019-10-10 2020-02-21 苏宁云计算有限公司 用于电商平台的智能调货方法及装置

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20070156883A1 (en) * 2005-12-30 2007-07-05 Visual Exchange Network, Llc System and method for group to group entertainment
CN107730091A (zh) * 2017-09-26 2018-02-23 沈阳工业大学 一种单周期二级物流服务供应链的协调运作方法
CN108171459A (zh) * 2017-12-29 2018-06-15 长春师范大学 基于博弈论的智能仓储优化方法
US20190347371A1 (en) * 2018-05-09 2019-11-14 Volvo Car Corporation Method and system for orchestrating multi-party services using semi-cooperative nash equilibrium based on artificial intelligence, neural network models,reinforcement learning and finite-state automata
US20200167722A1 (en) * 2018-11-27 2020-05-28 Zume Inc. Delivery of food items by aerial or ground drones to and from delivery vehicles
CN110378577A (zh) * 2019-07-02 2019-10-25 武汉理工大学 一种基于Supply-Hub的供应商管理库存运作方法
CN111199375A (zh) * 2020-01-13 2020-05-26 江苏佳利达国际物流股份有限公司 一种智能化物流运输系统
CN113393040A (zh) * 2021-06-18 2021-09-14 重庆邮电大学工业互联网研究院 一种基于博弈论的工业园区物流调度方法及系统

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
"Doctoral Dissertation", vol. 2012, 31 July 2012, NANJING UNIVERSITY OF SCIENCE AND TECHNOLOGY, CN, article LI, KUILIN: "Coordinated-mechanism Design for Decentraized Supply Chain Scheduling", pages: 1 - 58, XP009542403 *

Cited By (16)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116342197A (zh) * 2023-01-05 2023-06-27 上海朗晖慧科技术有限公司 一种结合地理位置信息的实时价格方案生成方法
CN116342197B (zh) * 2023-01-05 2023-11-14 上海朗晖慧科技术有限公司 一种结合地理位置信息的实时价格方案生成方法
CN116108550A (zh) * 2023-04-13 2023-05-12 湖北工业大学 一种基于bim的多式联运的动态优化方法及系统
CN116187895A (zh) * 2023-04-24 2023-05-30 北京中超伟业信息安全技术股份有限公司 一种智能仓储货流规划方法、系统及电子设备
CN116187895B (zh) * 2023-04-24 2023-07-18 北京中超伟业信息安全技术股份有限公司 一种智能仓储货流规划方法、系统及电子设备
CN116873431B (zh) * 2023-07-07 2024-02-06 湘南学院 一种基于岩板智能仓库的多重载agv储运方法
CN116873431A (zh) * 2023-07-07 2023-10-13 湘南学院 一种基于岩板智能仓库的多重载agv储运方法
CN116823124A (zh) * 2023-08-30 2023-09-29 四川普鑫物流自动化设备工程有限公司 一种集中式多层货架四向车调度系统及方法
CN116823124B (zh) * 2023-08-30 2023-11-24 四川普鑫物流自动化设备工程有限公司 一种集中式多层货架四向车调度系统及方法
CN117196262A (zh) * 2023-11-06 2023-12-08 中船凌久高科(武汉)有限公司 一种基于状态编码优化的测试场车辆与场景匹配调度方法
CN117196262B (zh) * 2023-11-06 2024-02-13 中船凌久高科(武汉)有限公司 一种基于状态编码优化的测试场车辆与场景匹配调度方法
CN117455341A (zh) * 2023-12-25 2024-01-26 华东交通大学 一种智能化车货匹配灵活调优方法
CN117455341B (zh) * 2023-12-25 2024-04-02 华东交通大学 一种智能化车货匹配灵活调优方法
CN117808596A (zh) * 2024-01-26 2024-04-02 国金证券股份有限公司 证券期货行业基于长短期记忆模块的大语言模型多Agent系统
CN117933882A (zh) * 2024-02-01 2024-04-26 山东曼索信息科技有限公司 一种基于物联网的智能物流系统
CN118313736A (zh) * 2024-06-07 2024-07-09 张家港大裕橡胶制品有限公司 用于橡胶手套生产的复合产线加工管理方法及系统

Also Published As

Publication number Publication date
CN113393040B (zh) 2023-04-07
CN113393040A (zh) 2021-09-14
US20240054444A1 (en) 2024-02-15

Similar Documents

Publication Publication Date Title
WO2022262469A1 (zh) 一种基于博弈论的工业园区物流调度方法及系统
Singh et al. A matheuristic for AGV scheduling with battery constraints
CN109214756B (zh) 整车物流调度方法及装置、存储介质、终端
CN112418497B (zh) 一种面向制造物联的物料配送路径优化方法
CN109345091B (zh) 基于蚁群算法的整车物流调度方法及装置、存储介质、终端
CN116187896B (zh) 绿色车辆路径问题求解方法、装置、计算机设备以及介质
CN112990590B (zh) 一种网络货运平台背景下的电商物流中转优化方法及系统
CN110210666B (zh) 基于车货匹配的智能推荐方法、系统和存储介质
CN114237222B (zh) 一种基于强化学习的取送货车辆路径规划方法
CN104537446B (zh) 二层带模糊随机时间窗口车辆路径优化方法
CN115759917A (zh) 一种基于改进混合蚁群算法的物流路径规划方法
CN113361073A (zh) 一种基于改进拉格朗日松弛算法的库存路径联合优化方法
CN115062868B (zh) 一种预聚类的车辆配送路径规划方法和装置
Oliskevych et al. SIMULATION OF CARGO DELIVERY BY ROAD CARRIER: CASE STUDY OF THE TRANSPORTATION COMPANY.
CN114239931B (zh) 基于改进蚁群算法实现物流仓储装车调度的方法及装置
CN116227773A (zh) 一种基于蚁群算法的配送路径优化方法
CN114757394A (zh) 基于工作量平衡的物流车辆路径优化方法、系统和介质
Wang et al. Research of oil product secondary distribution optimization based on collaborative distribution
CN111768133B (zh) 物品入库方法和装置
CN108022070B (zh) 一种混合装卸车辆协同调度运输方法
CN114663012A (zh) 一种基于区块链的生产物流管理系统及方法
Zhang et al. Location optimization of fresh food e-commerce front warehouse
Wang et al. [Retracted] Optimization Model of Logistics Task Allocation Based on Genetic Algorithm
CN108961012A (zh) 一种车主报价决策方法、设备及计算机存储介质
CN118278843A (zh) 一种基于深度强化学习的o2o外卖取送货路径规划方法

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22823952

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 22823952

Country of ref document: EP

Kind code of ref document: A1