Multi-stage coupling task distribution method and system based on network flow

Abstract

The invention discloses a multi-stage coupling task allocation method and system based on network flow, which are characterized in that a task allocation problem in supply chain management is described to a problem of minimum cost and maximum flow, a flow-cost network graph GU is established, four types of nodes and edges among different nodes are set to describe a task, and the task allocation problem in the supply chain management is described to the minimum cost and the maximum flow. According to the method, multi-stage coupling task distribution is realized, so that a task distribution model which is more in line with an operation rule of an actual complex system is constructed, a dynamic task environment can be described more accurately, when the constructed flow-cost network graph GU is solved, the decomposition of coupling constraint conditions is not carried out, the solution is directly carried out, the global optimization of the solution process is realized, and the calculation efficiency is improved. The tasks in the supply chain can be allocated more reasonably, so that the resource utilization rate is improved, and resource waste is avoided.

Description

Technical Field

[0001] The present invention belongs to the technical field of task allocation, and relates to a multi-stage coupled task allocation method and system based on network flow. Background Art

[0002] The multi-stage coupled task allocation problem widely exists in fields such as supply chain management and kill chain construction. Its core feature is that each decision-making stage is interrelated, and the decision in a single stage will have a direct impact on subsequent stages. Therefore, it is necessary to conduct systematic optimization from a global perspective. Taking supply chain management as an example, enterprises need to effectively organize suppliers, manufacturers, warehouses, distribution centers, and users, etc., to carry out product manufacturing, transportation, distribution, and sales management methods under the condition of meeting a certain customer service level. However, when traditional methods deal with such problems, it is difficult to effectively depict the coupling relationship between stages, resulting in insufficient global optimization effect. Specifically:

[0003] (I) Graph theory and network modeling methods

[0004] Researchers abstract the multi-stage task allocation as a path search or matching problem in a directed graph. By defining the attributes of nodes (tasks / resources) and edges (assignment relationships / transfer paths) (such as capacity, cost, time window), they use algorithms such as bipartite graph matching, minimum spanning tree, or shortest path to solve the local optimal solution within a stage. For example, in single-stage task allocation, the Hungarian algorithm can efficiently solve the minimum weight matching problem of a bipartite graph; but in a multi-stage scenario, due to the existence of coupling relationships between stages, it is necessary to construct a multi-layer graph model to connect the decisions of each stage, resulting in an exponential growth in the graph scale and a significant increase in computational complexity.

[0005] (II) Mathematical programming and dynamic programming methods

[0006] Based on the idea of dynamic programming, the multi-stage problem is decomposed into a sequence of sub-problems, and the dependency relationship between stages is transmitted through a state transition equation. For example, in the classic resource allocation problem, the state variable can be defined as the remaining resource amount, the decision variable is the resource allocation plan for the current stage, and the objective function is the sum of the costs of each stage. However, when the state space dimension is high, dynamic programming will face the "curse of dimensionality" and it is difficult to handle non-linear constraints and random disturbances. Therefore, researchers have proposed an integer programming model, which uses 0-1 variables to depict the assignment relationship between tasks and resources and introduces coupling constraints. But for large-scale problems, the solution time of integer programming grows exponentially with the number of variables and is only applicable to medium and small-scale scenarios.

[0007] (III) Intelligent optimization and heuristic algorithms

[0008] For large-scale complex problems, meta-heuristic algorithms such as genetic algorithms, particle swarm optimization, and ant colony algorithms are widely used. These algorithms perform heuristic searches in the solution space by simulating natural evolution or swarm intelligence behaviors and can obtain approximate optimal solutions within a reasonable time. However, the performance of intelligent algorithms highly depends on parameter settings (such as population size, number of iterations), and there is a problem of "premature convergence", making it difficult to guarantee the accuracy and stability of solutions. Especially in dynamic environments, they lack the ability of rapid re-optimization.

[0009] It can be seen that the problems existing in traditional methods are as follows:

[0010] (1) Incomplete modeling of coupling relationships

[0011] Traditional methods generally adopt a "decoupled optimization" strategy, that is, it is assumed that the inter-stage dependency relationships can be approximately decomposed into independent constraints, or a multi-stage problem is simplified to a cascaded solution of single-stage problems. This local-optimal solution idea ignores the two-way influence between stages and may lead to the problem of reduction of the global feasible solution space when the supply chain performs task allocation.

[0012] (2) The contradiction between computational complexity and solution accuracy

[0013] In large-scale, dynamic, and uncertain scenarios, traditional algorithms are difficult to balance computational complexity and solution quality, especially the rationality of resource allocation and the lack of system robustness. Summary of the Invention

[0014] The purpose of the present invention is to solve the problem in the prior art that when performing task allocation in supply chain management, the coupling relationship is decoupled and decomposed into single-stage problems for solution, which will inevitably cause the coupling relationship to be unclear, resulting in unreasonable task allocation in the supply chain and waste of resources, and to provide a multi-stage coupled task allocation method and system based on network flow.

[0015] To achieve the above purpose, the present invention adopts the following technical solutions:

[0016] A multi-stage coupled task allocation method based on network flow, comprising the following steps:

[0017] Construct a minimum-cost maximum-flow model for supply chain management, and determine a flow-cost network graph G describing supply chain management based on the minimum-cost maximum-flow model U ;

[0018] Solve the flow-cost network graph G of supply chain management U to obtain the task allocation result of supply chain management;

[0019] The flow-cost network graph G UIt includes edges between nodes, and the nodes include source nodes, sink nodes, customer nodes, path set nodes and service nodes. The edges between the nodes include edges between customer nodes and path set nodes, edges between source points and customer nodes, edges between path set nodes and service nodes, and edges between service nodes and sink nodes.

[0020] A further improvement of the present invention is:

[0021] The traffic-cost network graph G U There are Q customer nodes in the network, and the customer nodes are the traffic-cost network graph G U The first-level node in the definition of customer set is DD = {DD1, DD2, ..., DD Q}, where DD k (k∈{1,2,…,Q}) represents the process from the initiation of a customer's demand to the satisfaction of the demand.

[0022] The path set nodes represent all feasible paths from the supplier to the second-level forwarder via the first-level forwarder;

[0023] The path set nodes are the flow-cost network graph G U The secondary node in

[0024] Traffic-cost network graph G U There are R path set nodes, and the total number of feasible paths is R (R = n1n2n3), where n1 represents the number of manufacturers, n2 represents the number of wholesalers, and n3 represents the number of retailers.

[0025] The service nodes include all suppliers, first-tier forwarders and second-tier forwarders;

[0026] The service node is the traffic-cost network graph G U The third-level node in .

[0027] The edge between the client node and the path set node is defined as the total distance of the corresponding feasible path, denoted as D k , the flow on the edge between the client node and the path set node is 0 or 3;

[0028] Assume that the number of all service nodes is Num, Num = n1 + n2 + n3, the maximum capacity of the edge between the source and the customer node is the total number of service nodes Num, n1 represents the number of manufacturers, n2 represents the number of wholesalers; n3 represents the number of retailers

[0029] The capacity of the edge between the path set node and the service node is 1 and the cost is 0.

[0030] The edge between a service node and a sink has a capacity of 1 and a cost of 0.

[0031] A multi-stage coupled task allocation system based on network flow, comprising:

[0032] A model construction module, configured to construct a minimum-cost maximum-flow model for supply chain management, and obtain a flow-cost network graph G describing supply chain management U ;

[0033] A model solving module, configured to solve the flow-cost network graph G of supply chain management U to obtain a task allocation result of supply chain management;

[0034] A model description module, for the flow-cost network graph G U includes nodes and edges between nodes, the nodes include a source node, a sink node, a customer node, a path set node, and a service node, and the edges between the nodes include the edges between the customer node and the path set node, the edges between the source node and the customer node, the edges between the path set node and the service node, and the edges between the service node and the sink node.

[0035] A terminal device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the steps of any method of the present invention are implemented.

[0036] A computer-readable storage medium storing a computer program, wherein when the computer program is executed by a processor, the steps of any method of the present invention are implemented.

[0037] Compared with the prior art, the present invention has the following beneficial effects:

[0038] The present invention discloses a multi-stage coupled task allocation method based on network flow, which describes the task allocation problem in supply chain management as a minimum-cost maximum-flow problem. In the flow-cost network graph G U , four types of nodes are set, and the edges between different nodes are used to describe tasks, so as to realize multi-stage coupled task allocation, and to construct a task allocation model that better conforms to the operation rules of the actual complex system, so that it can more accurately depict the dynamic task environment. When the constructed flow-cost network graph G U is solved, the coupling constraint conditions are not decomposed, and the solution is directly carried out to achieve the global optimum of the solution process, which can more reasonably allocate tasks in the supply chain, thereby improving resource utilization rate and avoiding resource waste. Description of the Drawings

[0039] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and should not be regarded as limiting the scope. For those of ordinary skill in the art, without creative efforts, other related drawings can also be obtained based on these drawings.

[0040] Figure 1 It is a structural diagram of the multi-stage coupled task allocation problem abstracted.

[0041] Figure 2 It is the model structural diagram of the present invention. Detailed implementation manners

[0042] To make the objectives, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some but not all of the embodiments of the present invention. Usually, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.

[0043] Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the present invention claimed, but merely represents the selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0044] It should be noted that similar reference numerals and letters denote similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings.

[0045] In the description of the embodiments of the present invention, it should be noted that if terms such as "upper", "lower", "horizontal", "inner", etc. are used to indicate the orientation or positional relationship, it is based on the orientation or positional relationship shown in the drawings, or the orientation or positional relationship in which the product of the invention is usually placed when in use. It is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation of the present invention. In addition, terms such as "first", "second", etc. are only used for distinguishing descriptions and cannot be understood as indicating or implying relative importance.

[0046] In addition, if the term "horizontal" appears, it does not mean that the component is required to be absolutely horizontal, but it can be slightly inclined. For example, "horizontal" only means that its direction is more horizontal relative to "vertical", and does not mean that the structure must be completely horizontal, but it can be slightly inclined.

[0047] In the description of the embodiments of the present invention, it should also be noted that unless otherwise clearly specified and limited, if the terms "set", "installed", "connected", "connected" are used, they should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be directly connected or indirectly connected through an intermediate medium, and it can be the communication inside two components. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific situations.

[0048] The following further describes the present invention in detail with reference to the accompanying drawings:

[0049] See Figures 1 to 2 , the embodiments of the present invention disclose a multi-stage coupled task allocation method based on network flow. The present invention proposes to use the network flow theory as the core modeling tool to transform the multi-stage coupled task allocation problem abstracted from the supply chain management problem into a global modeling of the task allocation process by constructing a three-level flow-cost network including customer nodes, path set nodes, and service nodes. The natural advantages of the network flow model are:

[0050] (1) Explicit coupling relationship: directly characterize the dependencies between stages through the connection relationship, capacity, and cost attributes of the edges, avoiding information loss caused by decoupling;

[0051] (2) Efficient solution framework: use the mature minimum-cost maximum-flow algorithm to solve large-scale problems in polynomial time, balancing computational efficiency and accuracy;

[0052] (3) Domain generality: The model structure can be flexibly adjusted to adapt to different scenarios (such as adding or reducing node levels, modifying edge attributes), providing a unified framework for cross-domain applications.

[0053] The present invention abstracts the multi-stage coupled task allocation problem for typical scenarios in supply chain management, such as Figure 1 shown, where node T corresponds to the customer in the supply chain, node O corresponds to the retailer in the supply chain, node D corresponds to the wholesaler in the supply chain, and node A corresponds to the manufacturer in the supply chain; model the transfer relationship between the manufacturer, wholesaler, retailer, and customer as the edges between nodes. At this time, the decision-making problem in supply chain management becomes as Figure 1For the multi-stage coupled task allocation problem shown, the objective function of the supply chain management decision is: to minimize the transfer cost while meeting all customer demands. The constraints are: assuming that there is a one-to-one allocation from the manufacturer to the wholesaler, from the wholesaler to the retailer, and from the retailer to the customer, that is, one manufacturer can satisfy one wholesaler, one wholesaler can satisfy one retailer, and one retailer can satisfy one customer demand; in the multi-stage coupled task allocation problem, it means that node resources can only be assigned once to ensure that there are no resource conflicts in the solution of the multi-stage coupled task allocation problem.

[0054] To build a task allocation model that better conforms to the operating laws of actual complex systems, enabling it to more accurately depict the dynamic task environment; and through the refined general modeling methods and solution strategies, promote the popularization and application of the task-resource allocation model in different fields. Through the constructed network flow model and minimum cost flow solution algorithm, various resources can be allocated more reasonably, thereby improving resource utilization and avoiding resource waste.

[0055] Specifically, it includes the following steps:

[0056] Step 1: Build a minimum cost maximum flow model for supply chain management to obtain the flow-cost network diagram G that describes supply chain management U ;

[0057] Specifically:

[0058] Since the flow network of the minimum cost maximum flow model is a flow-cost network diagram, where each edge has two attributes: capacity and cost, the capacity of the edge represents the maximum flow that can pass through the edge, and the cost represents the cost per unit flow passing through the edge. In the embodiment of the present invention, the number of service nodes passed through is used as the flow, and the driving distance of a truck's one-time delivery task is used to represent the cost of the corresponding edge. Next is the detailed description of the flow-cost network diagram G U :

[0059] Graph G U The nodes in it are divided into four categories in total. One category is the auxiliary nodes added in the model, which are used as the source point and sink point of graph G U respectively. The other three categories of nodes are customer nodes, path set nodes, and service nodes. The detailed descriptions of these three categories of nodes are as follows:

[0060] Step 1.1: Customer nodes

[0061] The present invention takes meeting customer demands as the primary goal. Assuming there are Q customers in total, the present invention adds all customers to graph G U as the first-level nodes, that is, customer nodes, which are represented by blue dots, as Figure 2 shown.

[0062] Define the customer set as DD = {DD1, DD2, …, DD Q}, where DD k (k ∈ {1, 2, …, Q}) represents the process from the initiation of a customer's demand to the satisfaction of the demand.

[0063] Step 1.2: Path set nodes

[0064] In the present invention, to satisfy a customer's demand, the goods first depart from the manufacturer, pass through the wholesaler and the retailer twice, and then reach the customer. Therefore, in this scenario, the distribution task of the transporter needs to be executed three times. The combination from the manufacturer through the wholesaler to the retailer is denoted as a feasible path that can satisfy the customer's demand. List all the feasible paths and add them to the graph G U as secondary nodes, that is, path set nodes, which are represented by green dots, as Figure 2 shown.

[0065] Assume that the number of manufacturers in the scenario is n1, the number of wholesalers is n2, and the number of retailers is n3. Then the total number of feasible paths is R (R = n1n2n3), and there are R secondary nodes in the graph G U .

[0066] Step 1.3: Service nodes

[0067] The present invention regards all suppliers, wholesalers, and retailers as service nodes and adds them to the graph G U as tertiary nodes, which are represented by yellow dots, as Figure 2 shown. Then there are Num (Num = n1 + n2 + n3) service nodes in the graph G U .

[0068] Furthermore, the edge modeling of the flow - cost network graph G U in the embodiment of the present invention:

[0069] The invention uses different capacities and costs to represent different requirements of each edge, specifically:

[0070] Step 1.4: Edges between customer nodes and path set nodes

[0071] If a combination of a customer node and a path set node can form a supply chain, then there is an edge between the path set node and the customer node. The cost of this edge represents the total distance of the corresponding supply chain, and the distance is denoted as D k . In the present invention, a customer may correspond to multiple distribution routes of the transporter, that is, a customer node may have edges with multiple path set nodes.

[0072] The flow of the edge connecting the customer node and the path set node is either 0 or 3, and the flow cannot be adjusted step by step because we assume that customer demands are all met at once.

[0073] Step 1.5: The edge between the source node and the customer node

[0074] Each customer node may be connected to multiple task set nodes, so it is possible that each delivery task may be for the same customer. Given that the total number of all service nodes is Num, the maximum capacity of the edge between the source node and the customer node is the total number of service nodes Num, as Figure 2 shown.

[0075] Step 1.6: The edge between the path set node and the service node

[0076] The edge between the path set node and the service node represents the correspondence between the path set and the service node, that is, the path set node is only connected to the service nodes it contains. Therefore, the capacity of the edge between the path set node and the service node is 1, and the cost is 0.

[0077] Step 1.7: The edge between the service node and the sink node

[0078] The capacity of the edge between the service node and the sink node is set to 1, which means that the goods must pass through each service point at least once. Similarly, the cost of the edge between them is 0.

[0079] Step 2: Solve the flow-cost network graph G of the supply chain management U to obtain the task allocation result of the supply chain management.

[0080] The basic idea of finding the minimum-cost flow: Starting from the cost digraph N(f0) of the zero flow (f0 = 0), use the method of finding the shortest path to find the minimum-cost chain μ0 from v s to v1, and adjust the flow on μ0. So the new feasible flow f1 must be the minimum-cost feasible flow. If v(f1) = v, the calculation terminates. Otherwise, reconstruct the cost digraph N(f1) for the feasible flow f1, and continue to find the minimum-cost chain μ1 from v s to v1 in N(f1), and adjust the flow on μ1 to obtain the new feasible flow f2 which must also be the minimum-cost feasible flow... and so on until a feasible flow with flow v is found. If v(f) = v is already the flow of the maximum flow, then this minimum-cost flow is the minimum-cost maximum flow. Thus, the method of finding the minimum-cost flow is the combination of the method of finding the shortest path and the method of finding the maximum flow.

[0081] The specific steps are as follows:

[0082] First, start constructing the cost digraph N(f0) corresponding to f0 = 0.

[0083] Step 2.1: On N(f0), find the minimum-cost chain μ0 from v s to v t using the shortest path method, and adjust the flow on μ0 according to the formula of θ, f ij ', to obtain the flow digraph of the new feasible flow f1:

[0084]

[0085] Step 2.2: Reconstruct the cost digraph N(f1) corresponding to the feasible flow f1. Its vertex set is still V, and the arc set A is changed as follows based on observing the minimum-cost chain in the previous cost digraph N(f ° ):

[0086] Step 2.2.1: For each arc on the minimum-cost chain μ0:

[0087] If (v i , v j ) ∈ μ + , and f ij < c ij , or (v i , v j ) ∈ μ - , and f ij > 0, then add an arc with the opposite direction to the original arc and a weight of -b ij .

[0088] If (v i , v j ) ∈ μ + , and f ij = c ij , or (v i , v j ) ∈ μ - , and f ij = 0, then change the direction of the original arc and change the weight to -b ij .

[0089] Step 2.2.2: For each arc outside the minimum-cost chain μ0, its direction and weight remain unchanged.

[0090] Step 2.3: On N(f1), still use the shortest path method to find the minimum-cost chain μ1 from v s to v t . If the minimum-cost chain does not exist, then f1 is the minimum-cost maximum flow from v s to v t , and the calculation terminates. Otherwise, execute Step 2.4.

[0091] Step 2.4: On the minimum-cost chain μ1 in N(f1), still adjust the flow according to the formula of θ,f ij ', to obtain a new minimum-cost feasible flow f2. If v(f2) = v, the calculation terminates; otherwise, return to Step 2.2.

[0092] Furthermore, this embodiment also discloses a simulation experiment based on this method:

[0093] Construct a multi-stage coupled task assignment network model:

[0094] Taking the supply chain scenario as an example, assume there are 3 manufacturers [A1, A2, A3], the production volumes of the manufacturers are [100, 110, 120], and 9 wholesalers [D1, D2, D3,..., D9];

[0095] The wholesale volumes of the wholesalers are as follows:

[0096] [10, 20, 30, 40, 50, 60, 70, 80, 90], and 4 retailers [O1, O2, O3, O4];

[0097] The purchase volumes of the retailers are as follows:

[0098] [100, 150, 200, 250], and 10 demanders [T1, T2, T3,..., T9, T 10 ;

[0099] The demand volumes of the demanders are as follows:

[0100] [50, 60, 70, 80, 90, 100, 110, 120, 130, 140].

[0101] According to the network flow theory, construct a supply chain network model N * (V, E, C, W). Determine the node set V, which includes the source node, manufacturer nodes, wholesaler nodes, virtual wholesaler nodes, retailer nodes, virtual retailer nodes, demander nodes, and sink node. Construct the directed edge set E, and at the same time set the capacity C and cost W of the edges.

[0102] Transform this problem into a minimum-cost maximum-flow problem, and use the minimum-cost maximum-flow algorithm to solve it. Starting from the initial zero flow, continuously construct the residual network, find the shortest path from the source node to the sink node in the residual network, and increase the flow along the shortest path. For example, in a certain iteration, find a shortest path from the source node through O, D, A to T, and increase one unit of flow on this path. Repeat this process until an optimal supply chain solution is obtained and the optimal supply plan is determined.

[0103] The comparison algorithm selects the traditional supply chain construction method based on manual experience. Simulate the above local supply scenarios 100 times, and generate supply chain solutions using the algorithm of the present invention and the traditional algorithm respectively.

[0104] Comparison of resource utilization rate:

[0105] Algorithm of the present invention: In these 100 simulations, count the number of manufacturers, retailers, and wholesalers used each time. For example, on average, 4 manufacturers, 2.5 wholesalers, and 3.5 retailers are used each time. There are a total of 6 manufacturers, 3 wholesalers, and 5 retailers available, and the resource utilization rate is (4 + 2.5 + 3.5) ÷ (6 + 3 + 5) × 100% = 62.5%.

[0106] Traditional algorithm: Similarly, count the number of resources used. On average, 5 manufacturers, 2.8 wholesalers, and 4 retailers are used each time, and the resource utilization rate is (5 + 2.8 + 4) ÷ (6 + 3 + 5) × 100% = 73.75%. The algorithm of the present invention can more reasonably plan resources through the network flow model, reduce unnecessary resource investment on the premise of ensuring the completion of the supply task, and the advantage will be more obvious as the supply scale and complexity increase.

[0107] Comparison of robustness:

[0108] Simulate the interference scenario, randomly make 20% of the nodes fail, and then count the number of times the supply chains generated by the two algorithms can still complete the supply task. Algorithm of the present invention: In 100 simulated interference experiments, the supply chain can successfully complete the task 70 times, and the robustness is 70%. This benefits from the dynamic weight mechanism and redundant path optimization. When some nodes fail, the network flow will be redistributed to other available nodes. Traditional algorithm: In the same 100 simulated interference experiments, only 40 times can complete the supply task, and the robustness is 40%. Due to the lack of an effective response mechanism in the traditional algorithm, once a node fails, the supply chain is easily interrupted and cannot complete the supply task.

[0109] Through the above specific implementation methods, case analysis, and comparative experiments, it can be seen that in terms of resource utilization rate and robustness, the method disclosed in the present invention has obvious advantages over the traditional algorithm. See Table 1 for the simulation results:

[0110] Table 1 Method of the present invention

[0111]

[0112]

[0113] Table 2 Traditional method

[0114]

[0115] This embodiment also discloses a multi-stage coupled task allocation system based on network flow, including:

[0116] A model construction module, used to construct a minimum-cost maximum-flow model for supply chain management and obtain a flow-cost network graph G describing supply chain management U ;

[0117] A model solving module, used to solve the flow-cost network graph G of supply chain management U and obtain the task allocation result of supply chain management.

[0118] A model description module, for the flow-cost network graph G U including nodes and edges between nodes, the nodes including a source node, a sink node, customer nodes, path set nodes, and service nodes, and the edges between the nodes including edges between customer nodes and path set nodes, edges between the source node and customer nodes, edges between path set nodes and service nodes, and edges between service nodes and the sink node.

[0119] The method and system disclosed in this faming have the following advantages:

[0120] Coupling relationship modeling: Using network flow theory to directly describe the dynamic coupling relationship of multi-stage tasks, avoiding information loss caused by decoupling, and achieving global optimization.

[0121] Minimum-cost maximum-flow transformation: Transforming the complex multi-stage assignment problem into a classical network flow problem and using a mature algorithm framework to solve it, taking into account both computational efficiency and accuracy.

[0122] A schematic diagram of a terminal device provided by an embodiment of the present invention. The terminal device of this embodiment includes: a processor, a memory, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps in the above-mentioned method embodiments are implemented. Alternatively, when the processor executes the computer program, the functions of each module / unit in the above-mentioned device embodiments are implemented.

[0123] The computer program can be divided into one or more modules / units, and the one or more modules / units are stored in the memory and executed by the processor to complete the present invention.

[0124] The terminal device can be a computing device such as a desktop computer, a notebook, a palm computer, and a cloud server. The terminal device may include, but is not limited to, a processor and a memory.

[0125] The processor may be a Central Processing Unit (CPU), or may also be other general-purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field-Programmable Gate Arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc.

[0126] The memory can be used to store the computer program and / or module. By running or executing the computer program and / or module stored in the memory, and by invoking the data stored in the memory, the processor realizes various functions of the terminal device.

[0127] If the modules / units integrated in the terminal device are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, to implement all or part of the processes in the above-described embodiment methods of the present invention, it can also be completed by instructing relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by the processor, the steps of the above-described various method embodiments can be realized. Among them, the computer program includes computer program code, and the computer program code can be in the form of source code, object code, executable file, or some intermediate form, etc. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disc, computer memory, Read-Only Memory (ROM), Random Access Memory (RAM), electrical carrier signal, telecommunication signal, and software distribution medium, etc. It should be noted that the content included in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, the computer-readable medium does not include electrical carrier signals and telecommunication signals.

[0128] The above are only the preferred embodiments of the present invention, and are not used to limit the present invention. For those skilled in the art, the present invention can have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A multi-stage coupled task allocation method based on network flow, characterized in that comprising the following steps: Construct a minimum-cost maximum-flow model for supply chain management, and determine a flow-cost network graph G for describing supply chain management based on the minimum-cost maximum-flow model U ; Solve the flow-cost network diagram G of supply chain management U to obtain the task allocation result of supply chain management; The flow-cost network graph G U includes nodes and edges between the nodes. The nodes include a source node, a sink node, customer nodes, path set nodes, and service nodes. The edges between the nodes include the edges between customer nodes and path set nodes, the edges between the source node and customer nodes, the edges between path set nodes and service nodes, and the edges between service nodes and sink nodes.

2. The multi-stage coupled task allocation method based on network flow according to claim 1, wherein The flow-cost network graph G U includes Q customer nodes, and the customer nodes are the first-level nodes in the flow-cost network graph G U . Define the customer set as DD = {DD1, DD2, …, DD Q}, where DD k (k ∈ {1, 2, …, Q}) represents the process from the initiation of a customer's demand to the satisfaction of the demand.

3. A multi-stage coupled task allocation method based on network flow according to claim 1, characterized in that, The path set nodes represent all feasible paths from suppliers through first-level transshipment merchants to second-level transshipment merchants; The path set node is a secondary node in the flow-cost network graph G U ; Flow - cost network diagram G U There are R path - set nodes, and the total number of feasible paths is R (R = n1n2n3), where n1 represents the number of manufacturers, n2 represents the number of wholesalers, and n3 represents the number of retailers.

4. A multi-stage coupled task allocation method based on network flow according to claim 1, characterized in that The service nodes include all suppliers, first-level transshipment merchants, and second-level transshipment merchants; The service node is a third-level node in the traffic-cost network graph G U as shown in 5. A multi-stage coupled task allocation method based on network flow according to claim 1, characterized in that The edge between the customer node and the path set node is defined as the total distance of the corresponding feasible path, denoted as D k , and the flow on the edge between the customer node and the path set node is 0 or 3; Set the number of all service nodes to Num, Num = n1 + n2 + n3, the maximum capacity of the edge between the source node and the customer node is the total number of service nodes Num, n1 represents the number of manufacturers, n2 represents the number of wholesalers; n3 represents the number of retailers.

6. The multi-stage coupled task allocation method based on network flow according to claim 5, wherein The capacity of the edge between the path set node and the service node is 1 and the cost is 0.

7. A multi-stage coupled task allocation method based on network flow according to claim 6, characterized in that The capacity of the edge between the service node and the sink node is set to 1 and the cost is 0.

8. A multi-stage coupled task allocation system based on network flow, characterized in that, including: A model construction module, configured to construct a minimum-cost maximum-flow model for supply chain management and obtain a flow-cost network diagram G that describes supply chain management U ; A model solving module, which is used to solve the flow-cost network graph G of supply chain management U to obtain the task assignment result of supply chain management; A model description module for the traffic-cost network graph G U comprises nodes and edges between the nodes. The nodes include a source node, a sink node, customer nodes, path set nodes, and service nodes. The edges between the nodes include the edges between customer nodes and path set nodes, the edges between the source node and customer nodes, the edges between path set nodes and service nodes, and the edges between service nodes and the sink node.

9. A terminal device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1-7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the method according to any one of claims 1-7.

Images