Risk early warning method and system based on automobile service supply chain

Abstract

The invention relates to a risk early warning method and system based on an automobile service supply chain, and the method comprises the following steps: carrying out the topological network modeling of the automobile service supply chain based on the supply chain data, and obtaining a topological network structure of the automobile service supply chain; whether a supply chain risk point exists in the topological network structure or not is recognized through a preset recognition algorithm, if yes, risk influence analysis is conducted on the supply chain data through the supply chain risk point in the topological network structure, and a risk influence analysis result is obtained; performing risk scene classification on the risk influence analysis result to obtain a risk scene type; and performing risk early warning evaluation based on the risk scene type to obtain a risk early warning evaluation result, and formulating a risk relief scheme based on the risk scene type and the risk early warning evaluation result, thereby solving the problem that unsmooth communication among links of a supply chain and difficulty in effectively coping with emergencies due to information asymmetry.

Description

Technical Field

[0001] The present invention relates to the technical field of automobile services, and in particular to a risk early warning method and system based on an automobile service supply chain. Background Art

[0002] With the acceleration of global economic integration and the rapid development of information technology, the automotive service supply chain is facing unprecedented challenges. On the one hand, the uncertainty of the market environment has increased, such as fluctuations in raw material prices and frequent policy adjustments, which may directly or indirectly affect the stability of the supply chain; on the other hand, with the increasing diversification and personalization of consumer demand, automakers and service providers need to respond to market changes more flexibly and efficiently, which requires the supply chain to have higher transparency and agility. However, in actual operations, due to problems such as information asymmetry and imperfect coordination mechanisms, communication between various links in the supply chain is poor, decision-making efficiency is low, and it is difficult to effectively respond to emergencies.

[0003] In addition, the automotive service supply chain involves multiple complex links, from parts production, vehicle manufacturing to sales and after-sales service, and potential risk points may appear in each link. For example, suppliers may not be able to deliver on time due to natural disasters, equipment failures, etc., and the logistics process may also encounter unexpected situations such as traffic interruptions and cargo damage. These problems will not only increase the operating costs of enterprises, but may also lead to a decline in customer satisfaction and even damage brand reputation. Therefore, how to accurately identify and promptly deal with risks in the supply chain has become one of the problems that enterprises need to solve urgently.

[0004] In response to the above challenges, this study proposed a risk warning method based on the automotive service supply chain, which aims to comprehensively collect and analyze data from all links in the supply chain, use advanced graph theory algorithms to build a topological network model of the supply chain, and combine specific identification algorithms to discover potential risk points. Subsequently, through in-depth impact analysis and scenario classification of these risk points, scientific risk warning assessment results can be provided to decision makers, thereby helping them take effective mitigation measures in advance and reduce the impact of supply chain risks on the normal operation of the enterprise. This method can not only improve the risk resistance of the supply chain, but also promote the healthy development of the entire industry. Summary of the invention

[0005] The main purpose of the present invention is to provide a risk warning method and system based on the automotive service supply chain, which solves the problem of poor communication between various links in the supply chain due to information asymmetry and makes it difficult to effectively respond to emergencies.

[0006] To achieve the above object, the present invention provides a risk early warning method based on an automobile service supply chain, comprising the following steps: Collect data on each link in the automotive service supply chain to obtain supply chain data; Using a preset graph theory algorithm, based on the supply chain data, perform topological network modeling on the automotive service supply chain to obtain the topological network structure of the automotive service supply chain; Use a preset recognition algorithm to identify whether there are supply chain risk points in the topological network structure. If there are, perform risk impact analysis on the supply chain data through the supply chain risk points in the topological network structure to obtain a risk impact analysis result; Classify the risk impact analysis results into risk scenarios to obtain risk scenario types; Based on the risk scenario types, conduct risk early warning assessment to obtain a risk early warning assessment result, and formulate a risk mitigation plan based on the risk scenario types and the risk early warning assessment result.

[0007] Furthermore, each link includes suppliers, manufacturers, distributors, and retailers. The collection of data on each link in the automotive service supply chain to obtain supply chain data includes: Through a distributed crawler system, collect multi-source heterogeneous data from suppliers, manufacturers, distributors, and retailers in the automotive service supply chain to obtain original supply chain data; Preprocess the original supply chain data to obtain clean supply chain data; Through data fusion technology, fuse the clean supply chain data from different sources to obtain fusion data in a unified format; Parse the unstructured text data in the fusion data to extract key information in the fusion data; Use a preset long short-term memory network to predict the trend of the key information to obtain supply chain activity trend data, and use the supply chain activity trend data as supply chain data.

[0008] Furthermore, the use of a preset graph theory algorithm to perform topological network modeling on the automotive service supply chain based on the supply chain data to obtain the topological network structure of the automotive service supply chain includes: Using a preset graph theory algorithm, extract the entities and the relationships between the entities in the supply chain data; among them, the entities include suppliers, manufacturers, distributors, and retailers; Based on the entities and the relationships between the entities, construct a supply chain entity relationship table; among them, the supply chain entity relationship table includes nodes and edges, use the entities as nodes, and use the relationships between the entities as edges; Calculate the importance of each node in the supply chain entity relationship table using the node degree centrality layer in the graph theory algorithm to obtain a node importance score; wherein, the node importance score includes the in-degree and out-degree of each node. Calculate the frequencies of the in-degree and out-degree of each of the nodes to obtain the in-degree frequencies and out-degree frequencies of the nodes. Based on the node importance score, identify the corresponding entities and the upstream and downstream entities of the corresponding entities in the supply chain entity relationship table through a preset community detection algorithm to obtain a supply chain community structure diagram. Label the entities corresponding to the node importance score to obtain the labeled corresponding entities, and attach the labeled corresponding entities to the supply chain community structure diagram to obtain a labeled supply chain community structure diagram. Calculate the transmission paths of the edges between the nodes based on the in-degree frequencies and out-degree frequencies between the nodes; wherein, the transmission path represents the importance between the nodes, the shorter the transmission path, the closer the connection between the nodes, and the longer the transmission path, the looser the connection between the nodes. Based on the labeled supply chain community structure diagram and the transmission path, perform topological network modeling on the automotive service supply chain to obtain the topological network structure of the automotive service supply chain.

[0009] Further, perform risk impact analysis on the supply chain data through the supply chain risk points in the topological network structure to obtain a risk impact analysis result; wherein, the risk impact analysis result includes potential risk points and the influence scope of the supply chain risk points, including: Perform correlation analysis on the supply chain data through the supply chain risk points in the topological network structure to obtain a risk propagation path diagram. Based on the risk propagation path diagram, calculate the propagation probability of the supply chain risk points to the transmission path to obtain a risk propagation probability matrix; wherein, the rows of the risk propagation probability matrix represent the starting nodes of the transmission path, the columns of the risk propagation probability matrix represent the ending nodes of the transmission path, and the elements of the risk propagation probability matrix represent the propagation probability values from the starting node to the ending node. Extract the temporal characteristics of the risk propagation probability matrix to obtain a risk propagation temporal characteristic vector. Based on the risk propagation temporal characteristic vector, perform risk prediction on the supply chain data to obtain a risk impact degree prediction result, and use the risk impact degree prediction result as the risk impact analysis result. Based on the risk impact degree prediction result, perform vulnerability assessment on the supply chain data to obtain a network vulnerability index. When the network vulnerability index is less than the preset vulnerability index, determine the potential risk points based on the network vulnerability index; Conduct spatial analysis on the network vulnerability index to obtain the influence range of the supply chain risk points.

[0010] Furthermore, classify the risk scenario types for the risk impact analysis results, including: Use the potential risk points of the risk impact analysis results as rows and the influence range as columns to construct a risk feature matrix; wherein, the elements of the risk feature matrix represent the risk level; Input the risk feature matrix into a preset multi-level decision tree algorithm for analysis to obtain a risk feature hierarchy tree; Use the support vector machine classification algorithm to preliminarily classify the risk feature hierarchy tree to obtain a preliminary risk scenario type; Conduct risk scenario probability inference on the preliminary risk scenario type to obtain a probability risk scenario type; Conduct correlation analysis on the probability risk scenario type to obtain a correlation risk scenario type; Conduct risk scenario classification based on the preliminary risk scenario type, probability risk scenario type, and correlation risk scenario type to obtain the risk scenario type.

[0011] Furthermore, conduct risk early warning assessment based on the risk scenario type to obtain a risk early warning assessment result, and formulate a risk mitigation plan based on the risk scenario type and the risk early warning assessment result, including: Perform vector transformation on the risk scenario type to obtain a type transformation vector; Perform high-dimensional mapping on the type transformation vector to obtain an enhanced risk scenario feature vector; Perform adaptive weight allocation on the enhanced risk scenario feature vector to obtain a weighted risk feature vector; Perform time series analysis on the weighted risk feature vector through a preset long short-term memory network to obtain a risk early warning assessment result; Perform risk path simulation on the supply chain data based on the risk early warning assessment result to obtain a risk development path; Obtain a risk mitigation plan based on the risk development path and risk scenario type.

[0012] Furthermore, the risk mitigation plan includes a diversified supplier strategy plan, a multi-channel logistics strategy plan, and a resource reserve strategy plan.

[0013] The present invention also provides a risk early warning system based on an automotive service supply chain, including: The collection module is used to collect data from various links in the automotive service supply chain to obtain supply chain data; A modeling module, configured to use a preset graph theory algorithm to perform topological network modeling on the automobile service supply chain based on the supply chain data to obtain a topological network structure of the automobile service supply chain; An analysis module, used to identify whether there are supply chain risk points in the topological network structure through a preset identification algorithm, and if so, to perform risk impact analysis on the supply chain data through the supply chain risk points in the topological network structure to obtain risk impact analysis results; A classification module is used to classify the risk impact analysis results into risk scenarios to obtain risk scenario types; The assessment module is used to perform risk warning assessment based on the risk scenario type, obtain risk warning assessment results, and formulate risk mitigation plans based on the risk scenario type and the risk warning assessment results.

[0014] The present invention also provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of any one of the above methods when executing the computer program.

[0015] The present invention also provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps of any of the above-mentioned methods are implemented.

[0016] The risk warning method based on the automobile service supply chain provided by the present invention comprises the following steps: collecting data from each link in the automobile service supply chain to obtain supply chain data; using a preset graph theory algorithm to perform topological network modeling on the automobile service supply chain based on the supply chain data to obtain a topological network structure of the automobile service supply chain; identifying whether there is a supply chain risk point in the topological network structure by a preset recognition algorithm, and if so, performing risk impact analysis on the supply chain data through the supply chain risk point in the topological network structure to obtain a risk impact analysis result; classifying the risk impact analysis result by risk scenario to obtain a risk scenario type; performing risk warning assessment based on the risk scenario type to obtain a risk warning assessment result, and formulating a risk mitigation plan based on the risk scenario type and the risk warning assessment result. Through the above technical solution, the problem of poor communication between the various links of the supply chain due to information asymmetry is solved, and it is difficult to effectively respond to emergencies. It is realized that by classifying the risk impact analysis results by risk scenario, different types of risks can be classified and managed, which is convenient for enterprises to formulate targeted risk mitigation strategies according to different risk scenarios. This refined management method helps to improve the resilience and decision-making level of enterprises in the face of complex and changing market environments. Brief Description of the Drawings

[0017] Figure 1 is a schematic diagram of the steps of a risk early warning method based on an automotive service supply chain in an embodiment of the present invention; Figure 2 is a block diagram of the structure of a risk early warning system based on an automotive service supply chain in an embodiment of the present invention; Figure 3 is a schematic block diagram of the structure of a computer device in an embodiment of the present invention.

[0018] The realization of the object, functional features and advantages of the present invention will be further described with reference to the embodiments and the accompanying drawings. Detailed Embodiments

[0019] In order to make the object, technical solution and advantages of the present invention more clear and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0020] As Figure 1 shown Figure 1 is a schematic diagram of the steps of a risk early warning method based on an automotive service supply chain in an embodiment of the present invention; An embodiment of the present invention provides a risk early warning method based on an automotive service supply chain, including the following steps: Step S1, collect data of each link in the automotive service supply chain to obtain supply chain data.

[0021] Specifically, collecting data from all links in the automotive service supply chain to obtain supply chain data is the basis of the entire risk warning method. Specifically, to achieve this step, it is first necessary to determine the scope and objects of data collection, that is, to cover all key links of the supply chain from raw material procurement, parts production, vehicle assembly, logistics transportation to final sales and after-sales service. For example, in the process of automobile manufacturing, the quality reports of raw materials provided by suppliers, the time consumption records of each process on the production line, the inventory changes in the warehouse management system, and the customer feedback information at the sales terminal are all important data that need to be collected. By installing Internet of Things devices such as sensors and RFID tags, these data can be automatically and real-time collected to ensure the freshness and reliability of the data. At the same time, using big data platforms and technical means, data from different channels and in various formats can be integrated and cleaned to ensure the consistency and integrity of the data. In this way, not only can high-quality data support be provided for subsequent graph theory algorithm modeling and risk point identification, but it can also help enterprises comprehensively grasp the operating status of the supply chain, timely discover potential problems, and lay a solid foundation for achieving accurate risk warning. For example, in an actual application of supply chain risk warning, by continuously monitoring the quality data of parts provided by a certain automobile brand supplier, the system successfully predicted a large-scale recall risk that might be caused by unstable production of the supplier, thus saving a large amount of costs for the enterprise and avoiding reputation losses.

[0022] Step S2: Using a preset graph theory algorithm, based on the supply chain data, perform topological network modeling on the automotive service supply chain to obtain the topological network structure of the automotive service supply chain.

[0023] Specifically, using a preset graph theory algorithm, a topological network model of the automotive service supply chain is built based on the supply chain data, obtaining the topological network structure of the automotive service supply chain, which is one of the key steps in implementing the risk warning method. In this process, it is first necessary to select an appropriate graph theory algorithm, such as the shortest path algorithm, community detection algorithm, etc. These algorithms can help us understand and analyze the relationships and their complexities among the various nodes in the supply chain. Then, according to the previously collected supply chain data, each link in the supply chain is regarded as a node in the graph, and the interaction relationships such as logistics, information flow, and capital flow between the nodes are represented as edges. In this way, a topological network structure that can reflect the actual operation of the automotive service supply chain can be constructed. For example, in the automotive manufacturing industry, the vehicle manufacturer can be regarded as the central node in the network, around which there are many other nodes such as suppliers, distributors, and retailers. When these nodes are connected through various forms of interaction, a complex network is formed. In this network, not only can the position and role of each node be clearly seen, but also the stability of the supply chain and potential risk points can be evaluated by analyzing indicators such as the connection strength and path length between the nodes. For instance, if a key supplier has very close connections with multiple downstream enterprises, then once the supplier has problems, it may quickly affect many other nodes, leading to an impact on the entire supply chain. Through such modeling, enterprises can more intuitively understand the overall layout of the supply chain and provide strong support for subsequent risk identification and warning.

[0024] Step S3, identify whether there are supply chain risk points in the topological network structure through a preset identification algorithm. If there are, conduct a risk impact analysis on the supply chain data through the supply chain risk points in the topological network structure to obtain a risk impact analysis result.

[0025] Specifically, a pre-set recognition algorithm is used to identify whether there are supply chain risk points in the topological network structure. If there are, the supply chain risk points in the topological network structure are used to conduct a risk impact analysis on the supply chain data to obtain a risk impact analysis result. This process is the core link of the risk warning method. After completing the topological network modeling of the supply chain, the next step is to use a specially designed recognition algorithm to find potential risk points in the network. These recognition algorithms are usually based on machine learning or statistical principles and can automatically analyze the network structure features, such as the degree centrality and betweenness centrality of nodes, to determine which nodes or relationships between nodes may become the sources of risks. For example, in the automotive service supply chain, if the degree centrality of a certain supplier node is very high, it means that it has close connections with many other nodes. Then, once there are problems with this supplier, such as supply interruption or quality problems, it may have a wide impact on the entire supply chain. After identifying these risk points, the next step is to conduct an in-depth risk impact analysis on these risk points. This involves using the existing supply chain data to evaluate the possible impacts of the risk points on other parts of the supply chain, including but not limited to production delays, cost increases, and customer satisfaction decline. For example, assuming that the aforementioned supplier with high centrality has a serious quality problem, by analyzing data such as transaction records and historical performance between it and upstream and downstream nodes, the specific impacts such as the production delay time, additional quality control costs, and the number of customer complaints caused by this event can be predicted. Such analysis results can not only help enterprises understand the potential hazards of specific risk points but also provide an important basis for formulating effective risk mitigation strategies.

[0026] Step S4: Classify the risk impact analysis results into risk scenario types to obtain risk scenario types.

[0027] Specifically, classify the risk impact analysis results into risk scenario types. This step aims to further refine the results obtained from the risk impact analysis to facilitate subsequent risk early warning assessment and the formulation of mitigation plans. After completing the identification and impact analysis of supply chain risk points, we have obtained a preliminary understanding of the possible impacts of each risk point. However, these impacts may be multifaceted, involving different levels of the supply chain, such as production delays, cost increases, and decreased customer satisfaction. Therefore, in order to more effectively manage and respond to these risks, it is necessary to classify them according to certain criteria to form different risk scenario types. For example, in the automotive service supply chain, risk scenarios can be classified into several major categories such as supply chain disruptions, quality accidents, and changes in market demand. Each major category can be further divided into more sub-categories according to specific circumstances. For instance, supply chain disruptions can be further divided into critical raw material supply interruptions, important component supply interruptions, etc.; quality accidents can be divided into supplier product quality problems, quality problems during the manufacturing process, etc. Through such classification, enterprises can more clearly understand the specific impacts that different types of risks may have on their business, thereby providing a basis for formulating more targeted risk mitigation measures. For example, for the risk scenario of critical raw material supply interruption, enterprises can reduce risks by establishing a diversified supplier system, increasing safety inventory, etc.; for the risk scenario of supplier product quality problems, measures such as strengthening supplier quality management and introducing third-party quality certifications can be taken to prevent and reduce the occurrence of such risks. Such classification and refinement work not only helps enterprises enhance their understanding of risks but also promotes the optimization and improvement of their risk management strategies, thereby enhancing the resilience of the entire supply chain.

[0028] Step S5: Conduct a risk early warning assessment based on the risk scenario types to obtain a risk early warning assessment result, and formulate a risk mitigation plan based on the risk scenario types and the risk early warning assessment result.

[0029] Specifically, based on the risk scenario type, a risk early warning assessment is carried out to obtain the risk early warning assessment result. Based on the risk scenario type and the risk early warning assessment result, a risk mitigation plan is formulated. This process is the last and crucial step of the entire risk early warning method. After completing the risk scenario classification, the enterprise needs to conduct a detailed early warning assessment for each risk scenario to determine its likelihood of occurrence and potential impact level. In the assessment process, a combination of quantitative analysis and qualitative analysis can be adopted. Quantitative analysis mainly relies on historical data and statistical models to predict the probability of risk occurrence and the possible economic losses it may cause; qualitative analysis focuses on expert judgment and industry experience to evaluate the impact of risks on non-financial aspects such as brand image and customer satisfaction. For example, in the automotive service supply chain, if the risk scenario type is the interruption of key component supply, the enterprise can estimate the likelihood of such risks occurring in the future and the possible production delay days, additional cost expenditures, etc. by analyzing the occurrence frequency, duration, and impact on production plans of past similar events. At the same time, factors such as the credibility of suppliers and the availability of alternative suppliers also need to be considered to comprehensively evaluate the overall impact of the risk. After obtaining the risk early warning assessment result, the enterprise can formulate specific risk mitigation plans according to these results. These plans should propose practical preventive measures and emergency response plans according to the characteristics of different risk scenario types. Taking the interruption of key component supply as an example, the mitigation measures that the enterprise can take include, but are not limited to: establishing a more robust supply chain management system, establishing cooperative relationships with multiple suppliers to ensure the diversity of the supply chain; increasing safety inventory, especially for key components, to cope with emergencies; regularly conducting health checks on the supply chain to promptly discover and solve potential problems; strengthening communication and cooperation with suppliers to jointly improve the flexibility and response speed of the supply chain. In this way, not only can sufficient preparations be made before the risk occurs, but also rapid actions can be taken when the risk actually appears, minimizing losses to the greatest extent and ensuring the normal operation of the enterprise.

[0030] In a specific embodiment, each of the links includes suppliers, manufacturers, distributors, and retailers. Collecting data of each link in the automotive service supply chain to obtain supply chain data includes: Using a distributed crawler system to collect multi-source heterogeneous data from suppliers, manufacturers, distributors, and retailers in the automotive service supply chain to obtain the original supply chain data; Preprocessing the original supply chain data to obtain clean supply chain data; Using data fusion technology to fuse the clean supply chain data from different sources to obtain fused data in a unified format; Parsing the unstructured text data in the fused data to extract key information in the fused data; Predict the trend of the key information through a preset long short-term memory network to obtain supply chain activity trend data, and use the supply chain activity trend data as supply chain data.

[0031] Specifically, each of the links includes suppliers, manufacturers, distributors, and retailers. Collecting data on each link in the automotive service supply chain to obtain supply chain data is a complex and delicate process aimed at ensuring the comprehensiveness, accuracy, and availability of the data. First, a distributed crawler system is used to collect multi-source heterogeneous data from suppliers, manufacturers, distributors, and retailers in the automotive service supply chain to obtain the original supply chain data. This means using automated tools to scrape data from multiple platforms and systems on the Internet, and this data may come from the official websites of suppliers, the ERP systems of manufacturers, the sales records of distributors, and the customer feedback of retailers. For example, in the automotive manufacturing industry, material specifications and supply plans can be obtained from suppliers, production progress and quality inspection reports can be obtained from manufacturers, inventory levels and order status can be obtained from distributors, and sales performance and customer evaluations can be obtained from retailers. All of this data forms the basis of the original supply chain data. Next, the original supply chain data is preprocessed to obtain clean supply chain data. The preprocessing process mainly includes steps such as data cleaning, deduplication, and format conversion, with the aim of eliminating errors, redundancies, and inconsistencies in the data and bringing the data up to a certain quality and standard. For example, removing duplicate supplier records, correcting incorrectly formatted date fields, and filling in missing numerical values. Through this process, the accuracy and reliability of subsequent analysis can be ensured. Then, through data fusion technology, the clean supply chain data from different sources is fused to obtain fused data in a unified format. Data fusion refers to integrating data from different sources and different formats into a unified data set for comprehensive analysis. In the automotive service supply chain, this may involve matching and merging the bill of materials of suppliers with the production schedules of manufacturers, the inventory records of distributors with the sales data of retailers, etc. Data fusion technology can help enterprises overcome the challenges brought by multi-source data and achieve integrated management of data. Immediately following, data parsing is performed on the unstructured text data in the fused data to extract key information from the fused data. Unstructured text data refers to information that does not have a fixed format or structure, such as emails, social media comments, customer service records, etc. Through natural language processing (NLP) technology, valuable information can be extracted from it, such as customer satisfaction evaluations of a certain vehicle model, the credit ratings of suppliers, etc. These key information are crucial for understanding the dynamic changes in the supply chain. Finally, through a preset long short-term memory network (LSTM), trend prediction is performed on the key information to obtain supply chain activity trend data, and the supply chain activity trend data is used as the supply chain data. LSTM is a special type of recurrent neural network (RNN) that is particularly suitable for processing time series data and can capture the changing patterns of data over time.In the automotive service supply chain, historical data of key indicators such as the delivery time of suppliers, the production cycle of manufacturers, the inventory turnover rate of distributors, and the sales growth rate of retailers can be analyzed using an LSTM model to predict future trends. For example, if the model predicts that the demand for a certain model of car will increase significantly in the next quarter, the enterprise can increase inventory in advance to avoid sales losses caused by supply shortages. In this way, enterprises can better anticipate and respond to market changes, improving the flexibility and competitiveness of the supply chain.

[0032] In a specific embodiment, the use of a preset graph theory algorithm to perform topological network modeling on the automotive service supply chain based on the supply chain data to obtain the topological network structure of the automotive service supply chain includes: Using a preset graph theory algorithm, extract the entities in the supply chain data and the relationships between the entities; wherein, the entities include suppliers, manufacturers, distributors, and retailers; Construct a supply chain entity relationship table based on the entities and the relationships between the entities; wherein, the supply chain entity relationship table includes nodes and edges, using the entities as nodes and the relationships between the entities as edges; Use the node degree centrality layer in the graph theory algorithm to calculate the importance of each node in the supply chain entity relationship table to obtain a node importance score; wherein, the node importance score includes the in-degree and out-degree of each node; Calculate the frequencies of the in-degree and out-degree of each of the nodes to obtain the in-degree frequency and out-degree frequency of each node; Through a preset community detection algorithm, based on the node importance score, identify the corresponding entities and the upstream and downstream entities of the corresponding entities in the supply chain entity relationship table to obtain a supply chain community structure diagram; Label the entities corresponding to the node importance score to obtain the labeled corresponding entities, and attach the labeled corresponding entities to the supply chain community structure diagram to obtain a labeled supply chain community structure diagram; Calculate the transmission paths of the edges between each of the nodes based on the in-degree frequency and out-degree frequency between the nodes; wherein, the transmission path represents the importance between nodes, the shorter the transmission path, the closer the connection between nodes, and the longer the transmission path, the looser the connection between nodes; Perform topological network modeling on the automotive service supply chain based on the labeled supply chain community structure diagram and the transmission path to obtain the topological network structure of the automotive service supply chain.

[0033] Specifically, using a preset graph theory algorithm, a topological network model of the automotive service supply chain is constructed based on the supply chain data, obtaining the topological network structure of the automotive service supply chain. This is a multi-level and multi-step process aimed at revealing the complex relationships and structural characteristics within the supply chain through mathematical models and algorithms. First, using the preset graph theory algorithm, entities and the relationships between entities in the supply chain data are extracted. The entities mentioned here include suppliers, manufacturers, distributors, and retailers in the supply chain, and the relationships between entities cover the interactions of these entities in terms of materials, information, funds, etc. For example, in the automotive service supply chain, suppliers provide raw materials to manufacturers, manufacturers process these raw materials into automotive parts and then assemble them into complete vehicles, distributors are responsible for distributing the complete vehicles to each retailer, and retailers directly sell to consumers. These relationships constitute the basic framework of the supply chain. Next, a supply chain entity relationship table is constructed based on the entities and the relationships between entities. In this process, each entity in the supply chain is regarded as a node in the graph, and the relationships between entities are represented as edges between nodes. In this way, complex supply chain relationships can be simplified into an intuitive network model. For example, the supply relationship between supplier A and manufacturer B can be represented as a directed edge pointing from node A to node B. The supply chain entity relationship table not only contains the basic information of nodes and edges but may also contain the weights of the edges, such as transaction volume, transaction frequency, etc. These information helps with subsequent analysis. Then, using the node degree centrality layer in the graph theory algorithm, importance calculations are performed on each node in the supply chain entity relationship table to obtain node importance scores. Node degree centrality is an important indicator to measure the importance of a node in a network, which reflects the number of connections of the node with other nodes. In a directed graph, the degree centrality of a node can be further divided into in-degree and out-degree, representing the number of edges pointing to the node and the number of edges starting from the node respectively. By calculating the in-degree and out-degree of each node, the importance score of each node can be obtained. The higher the score, the more important the node's position in the network. For example, in the automotive service supply chain, if a certain supplier has a high out-degree, it means that it supplies raw materials to multiple manufacturers and is a key node in the supply chain. Further, calculate the frequencies of the in-degree and out-degree of each of the nodes to obtain the in-degree frequencies and out-degree frequencies of each node. Here, the frequency refers to the number of times occurring within a unit time. For example, the number of times a certain supplier supplies goods to multiple manufacturers per month. By calculating these frequencies, the activity level and influence of the nodes in the supply chain can be more accurately evaluated. For example, although a supplier has a high out-degree, if the supply frequency is low, it may not be as important as it seems on the surface. Through a preset community detection algorithm, based on the node importance scores, the corresponding entities and the upstream and downstream entities of the corresponding entities in the supply chain entity relationship table are identified to obtain the supply chain community structure diagram.Community detection algorithms can identify groups of closely connected nodes in a network, and these groups often represent different subsystems or sub-networks in a supply chain. For example, through community detection algorithms, it can be found that a close cooperation network has formed among certain suppliers and manufacturers, while another group of distributors and retailers constitutes another independent sub-network. After identifying these community structures, the internal operation mechanisms and external interaction patterns within each community can be analyzed in more depth. Label the entities corresponding to the importance scores of the nodes to obtain the corresponding labeled entities, and attach the labeled corresponding entities to the supply chain community structure diagram to obtain the labeled supply chain community structure diagram. This process not only increases the information content of the network model but also makes the model more intuitive and understandable. For example, in the supply chain community structure diagram, nodes with different importance levels can be distinguished by color or size, so it can be seen at a glance which nodes are the key nodes in the supply chain. Calculate the transmission paths of the edges between each pair of nodes based on the in-degree frequency and out-degree frequency between the nodes. Here, the transmission path represents the connection strength between nodes. The shorter the transmission path, the closer the connection between two nodes; conversely, the looser the connection. For example, in the automotive service supply chain, if a certain supplier has a high supply frequency with multiple manufacturers and a short supply time interval, it indicates that these manufacturers have a high degree of dependence on this supplier, and this path in the supply chain is relatively important. Finally, based on the labeled supply chain community structure diagram and the transmission paths, perform topological network modeling on the automotive service supply chain to obtain the topological network structure of the automotive service supply chain. Through this series of steps, not only can the overall architecture of the supply chain be intuitively displayed, but also the complex relationships and importance among the nodes in the supply chain can be revealed, providing a solid foundation for subsequent risk identification, impact analysis, and early warning assessment. For example, through the topological network structure, an enterprise can clearly see which suppliers are key nodes and which paths are the weak links in the supply chain, and thus take targeted measures to improve the stability and risk resistance of the supply chain.

[0034] In a specific embodiment, perform risk impact analysis on the supply chain data through the supply chain risk points in the topological network structure to obtain a risk impact analysis result; wherein, the risk impact analysis result includes potential risk points and the scope of influence of the supply chain risk points, including: Perform correlation analysis on the supply chain data through the supply chain risk points in the topological network structure to obtain a risk propagation path diagram; Based on the risk propagation path diagram, calculate the propagation probability of the supply chain risk points on the transmission path to obtain a risk propagation probability matrix; wherein, the rows of the risk propagation probability matrix represent the starting nodes of the transmission path, the columns of the risk propagation probability matrix represent the ending nodes of the transmission path, and the elements of the risk propagation probability matrix represent the propagation probability values from the starting node to the ending node; Extract the time series characteristics of the risk propagation probability matrix to obtain a risk propagation time series feature vector; Based on the risk propagation time series feature vector, conduct risk prediction on the supply chain data to obtain a risk impact degree prediction result, and use the risk impact degree prediction result as the risk impact analysis result; Conduct vulnerability assessment on the supply chain data based on the risk impact degree prediction result to obtain a network vulnerability index; When the network vulnerability index is less than the preset vulnerability index, determine the potential risk points based on the network vulnerability index; Conduct spatial analysis on the network vulnerability index to obtain the influence range of the supply chain risk points.

[0035] Specifically, the process of performing risk impact analysis on the supply chain data through the supply chain risk points in the topological network structure to obtain the risk impact analysis result is a systematic and meticulous process, aiming to comprehensively evaluate the scope and degree of influence of potential risk points in the supply chain. First, perform correlation analysis on the supply chain data through the supply chain risk points in the topological network structure to obtain a risk propagation path map. This means using the already constructed topological network structure, combined with the identified supply chain risk points, to analyze how these risk points spread to other parts through the nodes and edges in the network. For example, in the automotive service supply chain, if there is a supply problem with a certain key supplier, the risk propagation path map can be used to analyze how this problem affects the manufacturers, distributors, and retailers that rely on this supplier. The path map not only shows the path of risk propagation but also the speed and scope of propagation. Next, based on the risk propagation path map, calculate the propagation probability of the supply chain risk points on the transmission path to obtain a risk propagation probability matrix. The rows of the risk propagation probability matrix represent the starting nodes of the transmission path, the columns represent the ending nodes of the transmission path, and the elements of the matrix represent the propagation probability values from the starting node to the ending node. This process uses mathematical models and statistical methods to quantify the possibility of risk propagation in the supply chain network. For example, if the supply problem of Supplier A may affect Manufacturers B and C, calculate the propagation probability from A to B as 0.8 and from A to C as 0.6. These probability values form part of the risk propagation probability matrix. In this way, the path and possibility of risk propagation can be evaluated more accurately. Then, extract the temporal characteristics of the risk propagation probability matrix to obtain a risk propagation temporal characteristic vector. Temporal characteristic extraction refers to analyzing the probability change of risk propagation from the time dimension and extracting the key parameters that reflect the dynamic characteristics of risk propagation. For example, the change trend of the risk propagation probability over time can be calculated, or the fluctuations of risk propagation in different time periods can be analyzed. These temporal characteristic vectors can help us better understand the dynamic process of risk propagation and provide a basis for subsequent risk prediction. Based on the risk propagation temporal characteristic vector, perform risk prediction on the supply chain data to obtain a risk impact degree prediction result, and use the risk impact degree prediction result as the risk impact analysis result. This process uses machine learning or statistical models, combined with temporal characteristic vectors and historical data, to predict the impact degree of risk in a future period of time. For example, by analyzing the impact scope and duration of similar risk events in historical data, the specific impacts that the current risk event may cause to each link of the supply chain can be predicted, such as the number of production delay days and the increase in additional costs. These prediction results can not only help enterprises make preparations in advance but also provide a scientific basis for formulating risk mitigation measures. Based on the risk impact degree prediction result, perform vulnerability assessment on the supply chain data to obtain a network vulnerability index.The network vulnerability index is used to measure the vulnerability of the supply chain network when facing risks, and is usually calculated by comprehensively considering various factors such as network structure characteristics, node importance, and risk propagation probability. For example, if a node has a high in-degree and out-degree, and a high risk propagation probability, it indicates that the node is important in the network, and once a problem occurs, it may have a greater impact on the entire supply chain. By calculating the network vulnerability index, the weak links in the supply chain can be identified. When the network vulnerability index is less than the preset vulnerability index, the potential risk points are determined based on the network vulnerability index. The preset vulnerability index is a threshold used to judge whether the supply chain network is in a high-risk state. If the calculated network vulnerability index is lower than this threshold, it indicates that the supply chain network has a high risk and needs to be focused on and managed. For example, if the network vulnerability index of a certain supplier is much lower than the preset value, it indicates that the supplier is a potential risk point in the supply chain, and the enterprise needs to take measures to strengthen the cooperation management with the supplier to reduce the possibility of risks occurring. Conduct spatial analysis on the network vulnerability index to obtain the influence range of the supply chain risk points. Spatial analysis refers to analyzing the influence range of risk points on each link of the supply chain from the perspective of geographical space. For example, through map visualization tools, the influence degree of a certain risk point on suppliers, manufacturers, distributors, and retailers in different regions can be intuitively displayed. This kind of spatial analysis not only helps enterprises understand the overall impact of risks, but also provides support for formulating regional risk response strategies. For example, if a key supplier is located in an area with frequent natural disasters, the enterprise can increase the number of suppliers in other regions to disperse risks and improve the resilience of the supply chain. Through this series of steps, enterprises can not only comprehensively evaluate the influence range and degree of potential risk points in the supply chain, but also provide strong support for formulating scientific risk mitigation measures, thereby improving the stability and risk resistance ability of the supply chain.

[0036] In a specific embodiment, classifying the risk scenario of the risk impact analysis result to obtain risk scenario types, including: Taking the potential risk points of the risk impact analysis result as rows and the influence range as columns to construct a risk feature matrix; wherein, the elements of the risk feature matrix represent the risk degree; Inputting the risk feature matrix into a preset multi-level decision tree algorithm for analysis to obtain a risk feature hierarchy tree; Using a support vector machine classification algorithm to preliminarily classify the risk feature hierarchy tree to obtain a preliminary risk scenario type; Conducting risk scenario probability inference on the preliminary risk scenario type to obtain a probability risk scenario type; Conducting association analysis on the probability risk scenario type to obtain an associated risk scenario type; Based on the preliminary risk scenario type, the probabilistic risk scenario type, and the associated risk scenario type, perform risk scenario classification to obtain the risk scenario type.

[0037] Specifically, the process of classifying the risk impact analysis results into risk scenario types is a multi-stage and multi-level analysis process, aiming to transform the complex risk impact analysis results into risk scenario types that are easy to understand and manage. First, the potential risk points of the risk impact analysis results are used as rows, and the scope of influence is used as columns to construct a risk characteristic matrix. The risk characteristic matrix here is a two-dimensional table, where each row represents a potential risk point, each column represents the supply chain link or region that the risk point may affect, and the elements of the matrix represent the degree of risk. For example, in the automotive service supply chain, if it is identified that the supply problem of Supplier A is a potential risk point, and this problem may affect Manufacturers B, C, and D, then in the risk characteristic matrix, Supplier A is used as a row, and Manufacturers B, C, and D are used as columns respectively, and the elements in the matrix represent the degree of influence of Supplier A's supply problem on these three manufacturers, which can be a value between 0 and 1, where 1 represents the greatest influence and 0 represents no influence. Next, the risk characteristic matrix is input into a preset multi-level decision tree algorithm for analysis to obtain a risk characteristic hierarchy tree. The multi-level decision tree algorithm is a commonly used machine learning method that can perform multi-level partitioning and classification based on the characteristics of the input data. In this process, the algorithm will gradually divide different risk characteristic levels according to the data in the risk characteristic matrix. For example, the first level may be divided according to the type of risk point (such as supplier problems, logistics problems, etc.), and the second level may be further divided according to the scope of influence of the risk point (such as local influence, global influence, etc.). Through such hierarchical partitioning, the characteristics and influence paths of risk points can be displayed more clearly. Then, using the support vector machine classification algorithm, the risk characteristic hierarchy tree is preliminarily classified to obtain preliminary risk scenario types. Support Vector Machine (SVM) is a powerful classification algorithm that can handle high-dimensional data and find the optimal classification boundary. At this stage, by inputting the data in the risk characteristic hierarchy tree into the SVM algorithm, the risk points can be preliminarily classified into different risk scenario types. For example, the SVM algorithm may preliminarily classify Supplier A's supply problem into risk scenario types such as "supply chain interruption" or "quality accident". The results of the preliminary classification provide a basis for subsequent detailed analysis. Further, probability inference is performed on the preliminary risk scenario types to obtain probability risk scenario types. This process evaluates the probability of each preliminary risk scenario type through statistical methods or probability models. For example, it can be calculated that the probability of Supplier A's supply problem belonging to the "supply chain interruption" scenario is 0.7, and the probability of belonging to the "quality accident" scenario is 0.3. Through probability inference, the credibility of each risk scenario type can be evaluated more accurately, providing a basis for subsequent decision-making. Then, correlation analysis is performed on the probability risk scenario types to obtain correlated risk scenario types. Correlation analysis refers to discovering the relationships between different risk scenario types by mining the correlation relationships between data.For example, if the supply problems of Supplier A can lead to both "supply chain disruptions" and "quality accidents", there may be a certain correlation between these two risk scenario types. Through correlation analysis, compound risk scenarios can be identified, that is, a risk point may trigger multiple risk scenarios simultaneously. This analysis helps enterprises to more comprehensively understand the complexity of risks and formulate more comprehensive risk management strategies. Finally, based on the preliminary risk scenario types, probabilistic risk scenario types, and correlated risk scenario types, risk scenario classification is carried out to obtain risk scenario types. This final classification result comprehensively considers the analysis results at multiple levels, including not only the types of preliminary classification but also the results of probabilistic inference and correlation analysis. For example, the final risk scenario types may include "single supply chain disruption", "single quality accident", "compound supply chain disruption and quality accident", etc. Through such classification, enterprises can more accurately identify and manage different types of risks, providing a scientific basis for formulating specific countermeasures. Through this series of steps, enterprises can not only transform the results of complex risk impact analysis into risk scenario types that are easy to understand and manage but also more comprehensively evaluate and respond to various potential risks in the supply chain, thereby improving the stability and risk resistance of the supply chain. For example, in the automotive service supply chain, through the above method, enterprises can identify which suppliers are key risk points, which specific risk scenarios these risk points may trigger, and the scope and probability of these risk scenarios, so as to formulate targeted risk management strategies and improve the overall resilience of the supply chain.

[0038] In a specific embodiment, the risk early warning assessment is carried out based on the risk scenario types to obtain a risk early warning assessment result, and a risk mitigation plan is formulated based on the risk scenario types and the risk early warning assessment result, including: Perform vector transformation on the risk scenario types to obtain a type transformation vector; Perform high-dimensional mapping on the type transformation vector to obtain an enhanced risk scenario feature vector; Perform adaptive weight assignment on the enhanced risk scenario feature vector to obtain a weighted risk feature vector; Perform time series analysis on the weighted risk feature vector through a preset long short-term memory network to obtain a risk early warning assessment result; Perform risk path simulation on the supply chain data based on the risk early warning assessment result to obtain a risk development path; Obtain a risk mitigation plan based on the risk development path and the risk scenario types.

[0039] Specifically, the risk early warning assessment is carried out based on the risk scenario type to obtain the risk early warning assessment result, and a risk mitigation plan is formulated based on the risk scenario type and the risk early warning assessment result. This is a multi-level and multi-step process aimed at comprehensively evaluating and managing risks in the supply chain through advanced data analysis and machine learning techniques. First, the risk scenario type is vector-transformed to obtain a type transformation vector. This process transforms the qualitative risk scenario type into a quantitative vector form, facilitating subsequent mathematical operations and model processing. For example, in the automotive service supply chain, if the risk scenario types include "supply chain disruption", "quality accident", and "market demand change", these types can be respectively encoded as [1, 0, 0], [0, 1, 0], and [0, 0, 1] to form a type transformation vector. In this way, complex text descriptions can be transformed into digital forms that can be processed by computers. Next, the type transformation vector is mapped to a high-dimensional space to obtain an enhanced risk scenario feature vector. High-dimensional mapping is a technique that converts low-dimensional data into high-dimensional data, which can increase the expressive power and distinguishability of the data. For example, by using a kernel function (such as the radial basis function) for high-dimensional mapping, the above type transformation vector [1, 0, 0] can be mapped to a higher-dimensional vector, such as [1, 0, 0, 0.5, 0.3, 0.2]. The vector after high-dimensional mapping not only retains the original information but also adds new features, which helps to improve the accuracy and robustness of subsequent analysis. Then, an adaptive weight assignment is performed on the enhanced risk scenario feature vector to obtain a weighted risk feature vector. Adaptive weight assignment means dynamically adjusting the weight of each feature according to the importance of the feature to highlight key features and suppress noise. For example, by using an adaptive weight algorithm (such as gradient boosting or random forest), a weight value can be assigned to each feature in the enhanced risk scenario feature vector. These weight values reflect the importance of each feature in risk assessment and help to improve the prediction performance of the model. For example, if the "supply chain disruption" feature has a greater impact on risk in certain situations, its weight value may be set higher. Further, through a preset long short-term memory network (LSTM), time series analysis is performed on the weighted risk feature vector to obtain the risk early warning assessment result. LSTM is a special recurrent neural network that is particularly suitable for processing time series data and can capture the changing patterns of data over time. In this process, the LSTM model will predict the probability and impact degree of risk occurrence in the next period based on historical data and the current weighted risk feature vector. For example, by analyzing the occurrence frequency and impact scope of supply chain disruptions and quality accidents in the past few months, the LSTM model can predict the likelihood of these risk events occurring in the next month and their impact on the supply chain. This time series analysis can not only provide short-term risk early warnings but also help enterprises formulate long-term risk management strategies.Based on the risk early warning assessment results, simulate the risk path for the supply chain data to obtain the risk development path. Risk path simulation refers to using simulation technology to simulate the propagation process and development trend of risks in the supply chain. For example, if the LSTM model predicts that a certain key supplier may have supply problems within the next month, through risk path simulation, it can be analyzed how this problem affects the manufacturers, distributors, and retailers relying on this supplier, and how these impacts change over time. Through risk path simulation, enterprises can more intuitively understand the propagation path and influence scope of risks, providing a basis for formulating countermeasures. Finally, based on the risk development path and risk scenario types, obtain risk mitigation solutions. This process combines the results of risk early warning assessment and risk path simulation, formulating specific mitigation measures for each risk scenario type. For example, for the "supply chain interruption" risk scenario, enterprises can adopt measures such as establishing a diversified supplier system and increasing safety inventory; for the "quality accident" risk scenario, measures such as strengthening supplier quality management and introducing third-party quality certification can be taken. In this way, enterprises can not only make full preparations before the risk occurs but also take prompt actions when the risk actually appears, minimizing losses to the greatest extent and ensuring the normal operation of the supply chain. Through this series of steps, enterprises can not only comprehensively evaluate and manage various potential risks in the supply chain but also formulate scientific and effective risk mitigation solutions, improving the stability and risk resistance ability of the supply chain. For example, in the automotive service supply chain, through the above methods, enterprises can identify which suppliers are key risk points, what specific risk scenarios these risk points may trigger, as well as the influence scope and probability of these risk scenarios, so as to formulate targeted risk management strategies and improve the overall resilience of the supply chain.

[0040] In a specific embodiment, the risk mitigation solution includes a diversified supplier strategy solution, a multi-channel logistics strategy solution, and a resource reserve strategy solution.

[0041] The above described the risk early warning method based on the automotive service supply chain in the embodiments of the present invention. Next, the risk early warning system based on the automotive service supply chain in the embodiments of the present invention will be described. Please refer to Figure 2 , an embodiment of the risk early warning system based on the automotive service supply chain in the embodiments of the present invention includes: A collection module 21, configured to collect data of each link in the automotive service supply chain to obtain supply chain data; A modeling module 22, configured to use a preset graph theory algorithm to perform topological network modeling on the automotive service supply chain based on the supply chain data to obtain the topological network structure of the automotive service supply chain; An analysis module 23 is configured to identify whether there are supply chain risk points within the topological network structure through a preset recognition algorithm. If there are, risk impact analysis is performed on the supply chain data through the supply chain risk points in the topological network structure to obtain a risk impact analysis result. A classification module 24 is configured to classify the risk impact analysis result into risk scenarios to obtain risk scenario types. An evaluation module 25 is configured to perform risk early warning evaluation based on the risk scenario types to obtain a risk early warning evaluation result, and formulate a risk mitigation plan based on the risk scenario types and the risk early warning evaluation result.

[0042] In this embodiment, for the specific implementation of each unit in the above system embodiment, please refer to the description in the above method embodiment, and details are not described herein again.

[0043] Refer to Figure 3 , and this embodiment of the present invention also provides a computer device, the internal structure of which can be as Figure 3 shown. The computer device includes a processor, a memory, a display screen, an input device, a network interface, and a database connected through a system bus. Among them, the processor of the computer design is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store the corresponding data in this embodiment. The network interface of the computer device is used to communicate with an external terminal through a network connection. The computer program, when executed by the processor, implements the above method.

[0044] Those skilled in the art can understand that Figure 3 the structure shown in

[0045] is only a block diagram of a part of the structure related to the solution of the present invention, and does not constitute a limitation on the computer device to which the solution of the present invention is applied.

[0046] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, storage, database, or other medium provided in the present invention and used in the embodiments can include non-volatile and / or volatile memories. Non-volatile memories can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memories can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM can be obtained in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM, etc.

[0047] It should be noted that in this article, the term "including", "comprising", or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, device, article, or method including a series of elements not only includes those elements but also includes other elements not explicitly listed, or further includes elements inherent to such a process, device, article, or method. Without further limitation, an element defined by the statement "including one..." does not exclude the existence of another identical element in the process, device, article, or method including that element.

[0048] The above are only the preferred embodiments of the present invention, and do not limit the patent scope of the present invention accordingly. Any equivalent structural or equivalent process transformation made by using the description and drawings of the present invention, or directly or indirectly applied in other related technical fields, shall be equally included in the patent protection scope of the present invention.

Claims

1. A risk early warning method based on the automotive service supply chain, characterized in that, It includes the following steps: Collect data from all links in the automotive service supply chain to obtain supply chain data; Using a preset graph theory algorithm, based on the supply chain data, perform topological network modeling on the automotive service supply chain to obtain the topological network structure of the automotive service supply chain; Identify whether there are supply chain risk points in the topological network structure through a preset identification algorithm. If there are, perform risk impact analysis on the supply chain data through the supply chain risk points in the topological network structure to obtain a risk impact analysis result; Classify the risk impact analysis results into risk scenario types to obtain risk scenario types; Based on the risk scenario types, conduct risk early warning assessment to obtain a risk early warning assessment result, and formulate a risk mitigation plan based on the risk scenario types and the risk early warning assessment result.

2. The risk early warning method based on the automotive service supply chain according to claim 1, wherein The above-mentioned links include suppliers, manufacturers, distributors, and retailers. The collection of data from all links in the automotive service supply chain to obtain supply chain data includes: Use a distributed crawler system to collect multi-source heterogeneous data from suppliers, manufacturers, distributors, and retailers in the automotive service supply chain to obtain original supply chain data; Preprocess the original supply chain data to obtain clean supply chain data; Through data fusion technology, fuse the clean supply chain data from different sources to obtain fusion data in a unified format; Parse the unstructured text data in the fusion data to extract key information in the fusion data; Use a preset long short-term memory network to predict the trend of the key information to obtain supply chain activity trend data, and use the supply chain activity trend data as supply chain data.

3. The risk warning method based on the automotive service supply chain according to claim 2, wherein The use of a preset graph theory algorithm to perform topological network modeling on the automotive service supply chain based on the supply chain data to obtain the topological network structure of the automotive service supply chain includes: Use a preset graph theory algorithm to extract entities and relationships between entities in the supply chain data; among them, the entities include suppliers, manufacturers, distributors, and retailers; Construct a supply chain entity relationship table based on the entities and relationships between entities; among them, the supply chain entity relationship table includes nodes and edges, use the entities as nodes, and use the relationships between entities as edges; Use the node degree centrality layer in the graph theory algorithm to calculate the importance of each node in the supply chain entity relationship table to obtain a node importance score; among them, the node importance score includes the in-degree and out-degree of each node; Calculate the frequencies of the in-degree and out-degree of each node to obtain the in-degree frequency and out-degree frequency of each node; Through a preset community detection algorithm, based on the node importance score, identify the corresponding entities and the upstream and downstream entities of the corresponding entities in the supply chain entity relationship table to obtain a supply chain community structure diagram; Label the entities corresponding to the node importance score to obtain the labeled corresponding entities, and attach the labeled corresponding entities to the supply chain community structure diagram to obtain a labeled supply chain community structure diagram. Calculate the transmission paths of the edges between the nodes based on the in-degree frequency and out-degree frequency between the nodes; wherein, the transmission path represents the importance between the nodes, the shorter the transmission path, the closer the connection between the nodes, and the longer the transmission path, the looser the connection between the nodes; Based on the labeled supply chain community structure diagram and the transmission path, perform topological network modeling on the automotive service supply chain to obtain the topological network structure of the automotive service supply chain.

4. The risk warning method based on the automotive service supply chain according to claim 3, characterized in that Perform risk impact analysis on the supply chain data through the supply chain risk points in the topological network structure to obtain a risk impact analysis result; wherein, the risk impact analysis result includes potential risk points and the influence scope of the supply chain risk points, including: Perform correlation analysis on the supply chain data through the supply chain risk points in the topological network structure to obtain a risk propagation path diagram; Based on the risk propagation path diagram, calculate the propagation probability of the supply chain risk points to the transmission path to obtain a risk propagation probability matrix; wherein, the rows of the risk propagation probability matrix represent the starting nodes of the transmission path, the columns of the risk propagation probability matrix represent the ending nodes of the transmission path, and the elements of the risk propagation probability matrix represent the propagation probability values from the starting node to the ending node; Extract the time series characteristics of the risk propagation probability matrix to obtain a risk propagation time series feature vector; Based on the risk propagation time series feature vector, perform risk prediction on the supply chain data to obtain a risk impact degree prediction result, and use the risk impact degree prediction result as the risk impact analysis result; Based on the risk impact degree prediction result, perform vulnerability assessment on the supply chain data to obtain a network vulnerability index; When the network vulnerability index is less than the preset vulnerability index, determine the potential risk points based on the network vulnerability index; Perform spatial analysis on the network vulnerability index to obtain the influence scope of the supply chain risk points.

5. The risk early warning method based on the automotive service supply chain according to claim 4, characterized in that Perform risk scenario classification on the risk impact analysis result to obtain risk scenario types, including: Use the potential risk points of the risk impact analysis result as rows and the influence scope as columns to construct a risk feature matrix; wherein, the elements of the risk feature matrix represent the risk degree; Input the risk feature matrix into a preset multi-level decision tree algorithm for analysis to obtain a risk feature hierarchy tree; Use the support vector machine classification algorithm to perform preliminary classification on the risk feature hierarchy tree to obtain preliminary risk scenario types; Perform risk scenario probability inference on the preliminary risk scenario types to obtain probability risk scenario types; Perform correlation analysis on the probability risk scenario types to obtain associated risk scenario types; Based on the preliminary risk scenario types, probability risk scenario types and associated risk scenario types, perform risk scenario classification to obtain risk scenario types.

6. The risk warning method based on the automotive service supply chain according to claim 1, wherein Based on the risk scenario types, perform risk early warning assessment to obtain a risk early warning assessment result, and formulate a risk mitigation plan based on the risk scenario types and the risk early warning assessment result, including: Perform vector transformation on the risk scenario type to obtain a type transformation vector; Perform high-dimensional mapping on the type transformation vector to obtain an enhanced risk scenario feature vector; Perform adaptive weight allocation on the enhanced risk scenario feature vector to obtain a weighted risk feature vector; Perform time series analysis on the weighted risk feature vector through a preset long short-term memory network to obtain a risk warning assessment result; Based on the risk warning assessment result, perform risk path simulation on the supply chain data to obtain a risk development path; Based on the risk development path and the risk scenario type, obtain a risk mitigation plan.

7. The risk warning method based on the automotive service supply chain according to claim 1, characterized in that The risk mitigation plan includes a diversified supplier strategy plan, a multi-channel logistics strategy plan, and a resource reserve strategy plan.

8. A risk early warning system based on an automotive service supply chain, characterized in that, Comprising: A collection module for collecting data of each link in the automotive service supply chain to obtain supply chain data; A modeling module for using a preset graph theory algorithm to perform topological network modeling on the automotive service supply chain based on the supply chain data to obtain the topological network structure of the automotive service supply chain; An analysis module for identifying whether there are supply chain risk points in the topological network structure through a preset identification algorithm. If so, perform risk impact analysis on the supply chain data through the supply chain risk points in the topological network structure to obtain a risk impact analysis result; A classification module for performing risk scenario classification on the risk impact analysis result to obtain a risk scenario type; An evaluation module for performing risk warning assessment based on the risk scenario type to obtain a risk warning assessment result, and formulating a risk mitigation plan based on the risk scenario type and the risk warning assessment result.

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

10. A computer-readable storage medium having a computer program stored thereon, 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 to 7.

Images