Supply chain toughness evaluation method and device based on agent complex network

Abstract

The invention provides a supply chain toughness evaluation method and device based on an agent complex network, and the method comprises the steps: enabling each company node in a conventional supply chain network model to serve as an agent, enabling each agent to be provided with a perception sub-agent, a decision sub-agent and an execution sub-agent, enabling each node to be provided with the capability of autonomously perceiving, making a decision, and then carrying out the execution, company entities in a supply chain network are abstracted into intelligent agent models with perception, decision and execution capabilities, business exchange of the company entities is abstracted into connecting edges in a complex network, and through an interaction mechanism between intelligent agents, each company entity node in the supply chain network model has a dynamic characteristic; and the change of the operation decision of the company entity can be made according to the dynamic characteristics, so that a balance can be quickly achieved after disturbance, and the toughness evaluation can be quickly and accurately performed on the state of the supply chain network.

Description

Technical Field

[0001] This invention belongs to the field of complex networks, and in particular relates to a method and apparatus for assessing supply chain resilience based on agent-based complex networks. Background Technology

[0002] With the development of science and technology and economic globalization, supply chains have gradually evolved from traditional chain structures into complex network systems involving multiple levels of suppliers and manufacturers. The increasing complexity of supply chain network structures and disturbances in the external / internal environment (such as major public health events and corporate operational strategies) have exacerbated the vulnerability of supply chain networks, leading to frequent instances of systemic collapse caused by partial failures, resulting in significant economic losses. Therefore, it is necessary to conduct resilience assessments of supply chain networks and respond to network disturbances as early as possible.

[0003] Resilience, as one of the evaluation indicators of supply chain networks, is generally defined as the ability to maintain stable operation and recover rapidly in the face of various internal and external shocks and uncertainties. In existing literature, scholars have modeled supply chains as complex network models, treating corporate entities as "nodes" and business relationships between companies as "edges," and evaluating the resilience of supply chain networks under this model. However, because the nodes in existing supply chain network models are modeled as "two-state" models, with only two states—"intact" and "faulty"—they ignore the autonomous decision-making capabilities of corporate entities and the dynamic characteristics of their operations. This makes it impossible to describe the dynamic operations such as adjusting operational strategies adopted by companies to cope with crises, such as increasing inventory levels and strengthening cooperation with upstream and downstream enterprises, when conducting resilience assessments, thus failing to accurately evaluate the resilience level of the supply chain network. Summary of the Invention

[0004] The main problem addressed by this invention is how to conduct resilience assessment of supply chain networks and improve the accuracy of the assessment. It provides a supply chain resilience assessment method and apparatus based on intelligent agent complex networks.

[0005] To solve the above technical problems, the technical solution adopted is: A supply chain resilience assessment method based on intelligent agent complex network includes the following steps: Step 1: Obtain the corporate entities in the supply chain network and the business relationships between the corporate entities; Step 2: Construct a supply chain network model, with each company entity as a node, and establish directed connections between the corresponding nodes in the network model between companies with business relationships, using the business volume between companies as the weight of the connection between nodes; Step 3: Treat each node as an intelligent agent with autonomous decision-making capabilities, and each intelligent agent node has three sub-intelligent agent nodes that perform the functions of perception, decision-making and execution respectively. There is a data transmission relationship between the sub-intelligent agent nodes of the same intelligent agent node. Step 4: Each agent senses changes in the supply chain network model from itself or between agents. When a change occurs, it triggers the decision-making sub-agents of each agent node in the supply chain network model to make decisions and execute them, reaching a new stable node state. Step 5: Conduct a resilience assessment of the supply chain network model under the new steady state.

[0006] Furthermore, the method for triggering the decision-making sub-agents of each agent in the supply chain network model to make decisions and execute them is as follows: The perception sub-intelligent agents in each intelligent agent node perceive data of the external environment in the supply chain network. The decision-making sub-intelligent agents in each intelligent agent give corresponding operating strategies based on the data perceived by the perception sub-intelligent agents under the same intelligent agent. The execution sub-intelligent agents under the same intelligent agent adjust according to the operating strategies.

[0007] Furthermore, the method for the toughness assessment is as follows: toughness ; in This represents the number of nodes in the supply chain network that are operating normally under stable conditions. This represents the initial number of nodes in the supply chain network.

[0008] Furthermore, when the decision-making sub-agents of each agent node in the supply chain network model are triggered to make decisions and execute them, the agent nodes have two post-execution states: (1) Normal operation, when the intelligent agent node When the operating strategy is adjusted, the intelligent agent nodes connected to it... Then through its own perception-type sub-intelligent agent Perform perception and trigger decision-making sub-agents. and execution class sub-intelligent agents Adjust the current operating state of the intelligent agent node; (2) Unable to operate normally, when the agent node It senses the status of the intelligent agent nodes associated with it in the current supply chain network. When the fault threshold of its own intelligent agent node is triggered, it then shuts down the intelligent agent node. The status is changed to a fault status, and the current supply chain network model is exited.

[0009] Furthermore, during normal operation, when the perception-type sub-agent triggers the decision-type and execution-type sub-agents under the same agent node through perception, triggering means that several perception conditions are set in advance and a decision scheme is given under each perception condition. When a certain perception condition is reached, the decision-type sub-agent gives a decision scheme and hands it over to the execution-type sub-agent for execution.

[0010] Furthermore, one method to adjust the current operating status of an agent node is to adjust the connection relationship or workload between the current agent node and other agent nodes.

[0011] The present invention also provides a supply chain resilience assessment device based on intelligent agent complex network, comprising the following modules: acquisition module: used to acquire the corporate entities in the supply chain network and the business relationships between the corporate entities; Model building module: Used to build supply chain network model, with each company entity as a node, and to build directed connections between the corresponding nodes in the network model between companies with business relationships, and to use the business volume between companies as the weight of the connection relationship between nodes. Intelligent agent node construction module: used to treat each node as an intelligent agent with autonomous decision-making capabilities, and each intelligent agent node has three sub-intelligent agent nodes that perform the functions of perception, decision-making and execution respectively. There is a data transmission relationship between the sub-intelligent agent nodes of the same intelligent agent node. New steady-state module: used to perceive changes in the supply chain network model from the agent itself or between agents. When there is a change, it triggers the decision-making sub-agents of each agent node in the supply chain network model to make decisions and execute them, so as to reach a new steady state. Resilience Assessment Module: Used to assess the resilience of supply chain network models under new steady-state conditions.

[0012] By adopting the above technical solution, the present invention has the following beneficial effects: This invention provides a supply chain resilience assessment method and apparatus based on intelligent agent complex networks. By treating each company node in a traditional supply chain network model as an intelligent agent, and each intelligent agent possessing three sub-intelligent agents—perception, decision-making, and execution—each company node gains the ability to autonomously perceive, make decisions, and then execute. This changes the traditional supply chain network model where each company node only has two states: "intact" and "faulty." Supported by intelligent agent modeling technology, and considering the autonomous decision-making capabilities of company entities in the supply chain network in the real world environment, the company entities in the supply chain network are abstracted into intelligent agent models with perception, decision-making, and execution capabilities. Their business interactions are abstracted as edges in a complex network. Through the interaction mechanism between intelligent agents, each company entity node in the supply chain network model possesses dynamic characteristics and can make changes to the operational decisions of the company entity based on these dynamic characteristics. This allows for a rapid equilibrium to be reached after disturbances, enabling rapid resilience assessment of the supply chain network. Attached Figure Description

[0013] Figure 1 This is a system flowchart of the present invention. Detailed Implementation

[0014] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0015] Figure 1 This invention illustrates a specific embodiment of a supply chain resilience assessment method based on intelligent agent complex networks, comprising the following steps: Step 1: Obtain the company entities in the supply chain network and the business relationships between the company entities.

[0016] Step 2: Construct a supply chain network model, treating each company entity as a node. Establish directed connections between the corresponding nodes in the network model for companies with business relationships, using the business volume between company entities as the weight of the connections between nodes. In this embodiment, a directed graph is used. This represents a supply chain network model, where... This represents the set of nodes representing company entities in a supply chain network. , Represents the number of nodes; It represents the set of business relationships between company entities within a supply chain network. If the company and company If there is a business relationship between them, then ;on the contrary . This represents the weight of the connections between company entities in the supply chain network, such as supply volume and purchase volume.

[0017] Step 3: Each node is treated as an intelligent agent with autonomous decision-making capabilities. Each agent node has three sub-agent nodes: perception, decision-making, and execution, each fulfilling the perception, decision-making, and execution functions respectively. Data transfer relationships exist between the sub-agent nodes of the same agent node. In this embodiment, by abstracting each company node into an intelligent agent node, endowing it with perception, decision-making, and execution functions, a description of nodes with autonomous decision-making capabilities in a complex network is achieved. and Represents intelligent agent nodes The sub-agents are those that perceive, make decisions, and execute. Representing corporate entities Perception-type sub-intelligent agents, Representing corporate entities Decision-making sub-agents, Representing corporate entities The execution class of sub-agents is defined, and data transfer relationships between various sub-agents are created. Data transfer between sub-agents is achieved using... It indicates. Among them. This represents the data transfer from the perception-type sub-agent to the decision-type sub-agent. Data transfer from decision-making sub-agents to execution sub-agents. This represents the data transfer from the execution-type sub-agent to the decision-type sub-agent. If If the two types of sub-agents are in a data transfer relationship, then there is a data transfer relationship between them; otherwise... By transmitting data between sub-agents, it is possible to pass the perceived change data to the decision-making sub-agent for decision-making, and then the decision-making sub-agent passes the decision plan data to the execution sub-agent for execution.

[0018] Step 4: Each agent senses changes in the supply chain network model, either from itself or between agents. When a change occurs, it triggers the decision-making sub-agents of each agent node in the supply chain network model to make decisions and execute them, reaching a new stable state.

[0019] In this embodiment, the method for triggering the decision-making sub-agents of each agent node in the supply chain network model to make decisions and execute them is as follows: The perception sub-intelligent agents in each intelligent agent node perceive data of the external environment in the supply chain network. The decision-making sub-intelligent agents in each intelligent agent give corresponding operating strategies based on the data perceived by the perception sub-intelligent agents under the same intelligent agent. The execution sub-intelligent agents under the same intelligent agent adjust according to the operating strategies.

[0020] In this embodiment, when the decision-making sub-agents of each agent node in the supply chain network model are triggered to make decisions and execute them, the agent nodes have two post-execution states: (1) Normal operation, when the intelligent agent node When the operating strategy is adjusted, the intelligent agent nodes connected to it... Then through its own perception-type sub-intelligent agent Perform perception and trigger decision-making sub-agents. and execution class sub-intelligent agents The current operating state of the agent node is adjusted. In this embodiment, triggering refers to pre-setting several perception conditions and providing a decision-making scheme under each perception condition. When a certain perception condition is met, the decision-making sub-agent provides a decision-making scheme, which is then executed by the execution sub-agent. In this embodiment, the decision-making sub-agent pre-stores decision-making schemes corresponding to several perception conditions. Furthermore, when the perception sub-agent detects a change, it can notify the senior management of the company node, who will then manually provide a decision-making scheme for the execution sub-agent to execute. Normal operation generally requires adjusting the operating strategy based on changes. When the agent node... When the operating strategy is adjusted (i.e., the adjustment) (that is, changes in the external environment), and the intelligent agent nodes connected to it. Then it can be through its own perception-type sub-intelligent agent Perform perception and trigger decision-making sub-agents. and execution class sub-intelligent agents To adjust the current operational status of the company.

[0021] In this embodiment, for example, each sensing sub-agent perceives changes in "inter-company supply volume" through information such as contracts signed with other companies. After perceiving the change in the supply chain network model, the agent nodes adjust their operating strategies. For example, if company A stops supplying goods to company B (perceiving a change in the supply chain network model), the connection weight between company A and company B will decrease (if there is no supply at all, the weight drops to 0). Then, company B will purchase a larger quantity of goods from company C (i.e., adjust its operating strategy; for the supply chain network, this is equivalent to an increase in the connection weight between company B and company C). Moreover, company B is still operating normally at this time and has not gone bankrupt due to the "network change."

[0022] (2) Unable to operate normally, when the agent node It senses the status of the intelligent agent nodes associated with it in the current supply chain network. When the fault threshold of its own intelligent agent node is triggered, it then shuts down the intelligent agent node. The state is changed to a fault state, and the company exits the current supply chain network model. For example, in the initial state of the network, company B (i.e., agent node vi) purchases goods from companies A and C respectively. However, due to some reason, company A stops supplying goods to company B (i.e., the supply chain network has changed, and the perception sub-agents of company B's agent node sense the change in node B in the supply chain network), and company C is also unable to increase its supply intensity to company B. When company B's agent node decides to increase the supply intensity of C based on the corresponding changes, at this time, company B senses the states of other related nodes (i.e., company A and company C) and finds that the amount of goods supplied by company A and company C is insufficient to complete company B's own production and operation activities (i.e., company B's fault threshold has been reached). Therefore, company B cannot operate normally and exits the current supply chain network model.

[0023] In this embodiment, during normal operation, when a perception-type intelligent agent triggers decision-making and execution-type sub-intelligent agents under the same intelligent agent node through perception, the method for adjusting the operating state of the intelligent agent node is to adjust the connection relationship or business volume between the current intelligent agent node and other intelligent agent nodes, such as increasing inventory or enhancing cooperation between upstream and downstream enterprises. This will also involve other company entities in the supply chain network, i.e., other intelligent agent nodes.

[0024] This embodiment enables intelligent nodes in the supply chain network model to possess perception, decision-making, and execution sub-intelligent agents, thereby allowing them to respond to dynamic changes in the supply chain network and quickly reach a new equilibrium state, thus enabling a correct assessment of the resilience of the supply chain network.

[0025] Step 5: Conduct a resilience assessment of the supply chain network model under the new steady state.

[0026] In this embodiment, the method for toughness assessment is as follows: toughness ; in This represents the number of nodes in the supply chain network that are operating normally under stable conditions. This represents the initial number of nodes in the supply chain network.

[0027] This embodiment utilizes complex agent networks, considering the autonomous decision-making capabilities of company entities in supply chain networks in the real world. It abstracts these entities into agent models with perception, decision-making, and execution capabilities, and abstracts their business interactions as edges in a complex network. Through the interaction mechanism between agents, it realizes changes in the operational decisions of company entities in the supply chain network, thus breaking the traditional binary state limitation of supply chain network nodes ("intact" and "faulty") and effectively representing changes in the operational state of company entity nodes. Based on this, a supply chain network resilience assessment algorithm is proposed, applicable to the technical fields of resilience assessment of complex supply chain networks. When the supply chain network environment changes, compared to the "passive" changes of nodes in traditional supply chain networks, this invention, by constructing perception, decision-making, and execution-type sub-agents, can effectively represent the autonomous decisions of company entities in real-world supply chain networks, and thus correctly evaluate the resilience of the supply chain network.

[0028] The present invention also provides a supply chain resilience assessment device based on intelligent agent complex network, comprising the following modules: acquisition module: used to acquire the corporate entities in the supply chain network and the business relationships between the corporate entities; Model building module: Used to build supply chain network model, with each company entity as a node, and to build directed connections between the corresponding nodes in the network model between companies with business relationships, and to use the business volume between companies as the weight of the connection relationship between nodes. Intelligent agent node construction module: used to treat each node as an intelligent agent with autonomous decision-making capabilities, and each intelligent agent node has three sub-intelligent agent nodes that perform the functions of perception, decision-making and execution respectively. There is a data transmission relationship between the sub-intelligent agent nodes of the same intelligent agent node. New steady-state module: used to perceive changes in the supply chain network model from the agent itself or between agents. When there is a change, it triggers the decision-making sub-agents of each agent node in the supply chain network model to make decisions and execute them, so as to reach a new steady state. Resilience Assessment Module: Used to assess the resilience of supply chain network models under new steady-state conditions.

[0029] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A supply chain resilience assessment method based on agent-based complex networks, characterized in that, The steps include: Step 1: Obtain the corporate entities in the supply chain network, as well as the business relationships between these entities; Step 2: Construct a supply chain network model, treating each company entity as a node, and establish directed connections between the corresponding nodes in the network model for companies with business relationships. Use the business volume between company entities as the weight of the connection relationship between nodes. Step 3: Treat each node as an intelligent agent with autonomous decision-making capabilities, and each intelligent agent node has three sub-intelligent agent nodes that perform the functions of perception, decision-making and execution respectively. There is a data transmission relationship between the sub-intelligent agent nodes of the same intelligent agent node. Step 4: Each agent senses changes in the supply chain network model from itself or between agents. When a change occurs, it triggers the decision-making sub-agents of each agent node in the supply chain network model to make decisions and execute them, reaching a new stable state. Step 5: Conduct a resilience assessment of the supply chain network model under the new steady state.

2. The evaluation method according to claim 1, characterized in that, The method to trigger the decision-making sub-agents of each agent node in the supply chain network model to make and execute decisions is as follows: The perception sub-agents in each agent node perceive data of the external environment in the supply chain network. The decision-making sub-agents in each agent node give corresponding operating strategies based on the data perceived by the perception sub-agents under the same agent. The execution sub-agents under the same agent node adjust according to the operating strategies.

3. The evaluation method according to claim 1, characterized in that, The method for toughness assessment is as follows: toughness ; in This represents the number of nodes in the supply chain network that are operating normally under stable conditions. This represents the initial number of nodes in the supply chain network.

4. The evaluation method according to claim 2, characterized in that, When the decision-making sub-agents of each agent node in the supply chain network model make decisions and execute them, the agent nodes have two post-execution states: (1) Normal operation, when the intelligent agent node When the operating strategy is adjusted, the intelligent agent nodes connected to it... Then through its own perception-type sub-intelligent agent Perform perception and trigger decision-making sub-agents. and execution class sub-intelligent agents Adjust the current operating state of the intelligent agent node; (2) Unable to operate normally, when the agent node It senses the status of the intelligent agent nodes associated with it in the current supply chain network. When the fault threshold of its own intelligent agent node is triggered, it then shuts down the intelligent agent node. The status is changed to a fault status, and the current supply chain network model is exited.

5. The evaluation method according to claim 4, characterized in that, During normal operation, when a perception-type sub-agent triggers decision-type and execution-type sub-agents under the same agent node through perception, triggering means that several perception conditions are set in advance and a decision scheme is given under each perception condition. When a certain perception condition is met, the decision-type sub-agent gives a decision scheme and hands it over to the execution-type sub-agent for execution.

6. The evaluation method according to claim 4, characterized in that, The method to adjust the current operating status of an agent node is to adjust the connection relationship or workload between the current agent node and other agent nodes.

7. A supply chain resilience assessment device based on intelligent agent complex networks, characterized in that, It includes the following modules: Acquisition module: used to acquire company entities in the supply chain network, as well as the business relationships between these company entities; Model building module: Used to build supply chain network model, with each company entity as a node, and to build directed connections between the corresponding nodes in the network model between companies with business relationships, and to use the business volume between companies as the weight of the connection relationship between nodes. Intelligent agent node construction module: used to treat each node as an intelligent agent with autonomous decision-making capabilities, and each intelligent agent node has three sub-intelligent agent nodes that perform the functions of perception, decision-making and execution respectively. There is a data transmission relationship between the sub-intelligent agent nodes of the same intelligent agent node; New steady-state module: Used to perceive changes in the supply chain network model from the agent itself or between agents. When there is a change, it triggers the decision-making sub-agents of each agent node in the supply chain network model to make decisions and execute them, so as to reach a new steady state. Resilience Assessment Module: Used to assess the resilience of supply chain network models under new steady-state conditions.

Images