Supplier management method and system based on risk assessment

Abstract

The invention belongs to the technical field of data processing, and particularly relates to a supplier management method and system based on risk assessment, and the method comprises the steps: obtaining the risk indexes of a supplier according to the quality risk, delivery risk, financial risk and compliance risk of the supplier; according to the upstream and downstream relationship between the suppliers, constructing a risk propagation directed graph, and further obtaining a risk propagation index between any two suppliers; further introducing an ant colony algorithm to construct a purchase path, and in each iteration process, comprehensively considering self risk indexes of selectable suppliers, risk propagation indexes from existing nodes in a current path to the selectable suppliers, and risk and cost-oriented pheromone distribution in a historical path, so as to dynamically calculate the selection probability of each selectable supplier; and the ants are guided to gradually construct an optimal path meeting purchasing requirements. According to the invention, the objectivity and accuracy of risk identification are improved, and the early warning capability of enterprises for systematic risks is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing. More specifically, the present invention relates to a supplier management method and system based on risk assessment. Background Art

[0002] At present, when enterprises conduct supplier management and procurement path selection, they generally rely on empirical judgment or simple scoring mechanisms, lacking the systematic assessment and dynamic optimization capabilities for the overall risks of the supply chain. Traditional methods usually make decisions based on a single dimension (such as price, delivery cycle, or quality pass rate), ignoring the complex dependencies among suppliers and the propagation characteristics of risks in the supply chain, resulting in procurement strategies being difficult to cope with sudden risk events, such as the discontinuation of supply by key suppliers, the spread of quality problems, and financial crises.

[0003] In addition, in the prior art, the assessment of supplier risks mostly adopts static scoring methods, failing to conduct dynamic modeling by combining the upstream and downstream impacts, thus ignoring some "invisible high-risk nodes" that seemingly have low risks but have high propagation capabilities. Once an abnormality occurs, it may cause the entire supply chain to break down.

[0004] Therefore, there is an urgent need for an intelligent path construction mechanism that can integrate multi-dimensional risk indicators and cost control objectives to enhance the resilience and sustainability of the supply chain. Summary of the Invention

[0005] To solve the technical problem that the above prior art ignores the propagation characteristics of risks, resulting in procurement strategies being difficult to cope with sudden risk events, the present invention provides solutions in the following aspects.

[0006] In a first aspect, the present invention provides a supplier management method based on risk assessment, including: Obtaining the own risk indicators of a supplier according to the quality risk, delivery risk, financial risk, and compliance risk of the supplier; constructing a directed risk propagation graph according to the upstream and downstream relationships among the suppliers, and obtaining the risk propagation index between any two suppliers according to the directed risk propagation graph; Construct a procurement path using the ant colony algorithm. In each round of iteration of the ant colony algorithm, determine the heuristic factor of the optional supplier according to the own risk index of the optional supplier and the risk propagation index from the supplier corresponding to each node in the current path to the optional supplier; determine the risk-oriented pheromone of the edge between the supplier corresponding to the current node in the current path and the optional supplier according to the sum of the own risk coefficients of the suppliers included in each path obtained in the previous round of iteration; determine the cost-oriented pheromone of the edge between the supplier corresponding to the current node in the current path and the optional supplier according to the sum of the unit procurement costs of the suppliers included in each path obtained in the previous round of iteration; determine the selection probability of the optional supplier according to the risk-oriented pheromone, cost-oriented pheromone and heuristic factor; select a supplier from all optional suppliers as the next node of the current path according to the selection probability.

[0007] Preferably, the method for obtaining the own risk index of the supplier is as follows: Obtain the quality risk of the supplier according to the product qualification rate of the supplier; obtain the delivery risk of the supplier according to the on-time delivery rate of the supplier, the number of delayed deliveries and the delay time of each delayed delivery; obtain the financial risk of the supplier according to the financial status of the supplier; obtain the compliance risk of the supplier according to the number of violations of the supplier; perform weighted summation on the quality risk, delivery risk, financial risk and compliance risk to obtain the own risk index of the supplier.

[0008] Preferably, the construction of the risk propagation digraph includes: Take each supplier as a node respectively. For any two suppliers and , in response to supplier being the upstream supplier of supplier , establish a directed edge pointing from supplier to supplier , and determine the weight of the directed edge pointing from supplier to supplier : , where represents the quantity of the th product supplied by supplier to supplier ; represents the total quantity of the th product supplied by all upstream suppliers of supplier to supplier ; represents the quantity of the th product supplied by supplier to supplier The number of product types; Represents the maximum value function; A directed graph is formed based on all nodes, directed edges, and the weights of the directed edges, denoted as the risk propagation directed graph.

[0009] Preferably, the method for obtaining the risk propagation index is as follows: For any two suppliers and , obtain all paths from supplier to supplier in the risk propagation directed graph, and each path is regarded as a risk propagation path; The risk propagation index from supplier to supplier satisfies the expression: , , represents the self - risk index of supplier ; represents the weight of the th directed edge in the th risk propagation path from supplier to supplier in the risk propagation directed graph; represents the number of directed edges in the th risk propagation path from supplier to supplier in the risk propagation directed graph; represents the number of risk propagation paths from supplier to supplier in the risk propagation directed graph.

[0010] Preferably, the heuristic factor satisfies the expression: ; where, represents the heuristic factor of the th alternative supplier, represents the self - risk index of the th alternative supplier; represents the risk propagation index from the supplier corresponding to the rd node in the current path to the th alternative supplier; represents the number of existing nodes in the current path; represents the exponential function with the natural constant as the base.

[0011] Preferably, the risk - oriented pheromone satisfies the expression: ; Denote the current iteration as the In the k-th iteration, the supplier corresponding to the current node in the current path is denoted as supplier , where denotes the risk-oriented pheromone of the edge between the k-th iteration and the m-th to the n-th alternative suppliers; denotes the risk-oriented pheromone of the edge between the k-th iteration and the p-th to the q-th alternative suppliers; denotes the pheromone evaporation coefficient; denotes the increment of the risk-oriented pheromone of the edge between the k-th iteration and the r-th to the s-th alternative suppliers: , denotes the number of paths containing the edge between the k-th iteration from the u-th to the v-th alternative suppliers in the result obtained in the k-th iteration, denotes the total pheromone; denotes the sum of the self-risk coefficients of the suppliers corresponding to all nodes in the k-th iteration in the w-th to the x-th path containing the edge between the y-th to the

[0012] Preferably, the cost-oriented pheromone satisfies the expression: ; Denote the current iteration as the k-th iteration, and the supplier corresponding to the current node in the current path is denoted as supplier , where denotes the cost-oriented pheromone of the edge between the k-th iteration and the m-th to the n-th alternative suppliers; denotes the cost-oriented pheromone of the edge between the k-th iteration and the p-th to the q-th alternative suppliers; denotes the pheromone evaporation coefficient; denotes the increment of the cost-oriented pheromone of the edge between the k-th iteration and the r-th to the s-th alternative suppliers: , represents the number of paths containing the edges between the -th and the -th alternative suppliers in the result obtained from the -th iteration; represents the total amount of pheromone; represents the sum of the unit procurement costs of the suppliers corresponding to all the nodes in the -th path containing the edges between the -th and the -th alternative suppliers in the result obtained from the -th iteration.

[0013] Preferably, the selection probability of the alternative suppliers satisfies the expression: ; Denote the current iteration as the -th iteration, and denote the supplier corresponding to the current node in the current path as supplier . In the formula, represents the selection probability of the -th alternative supplier as the next node of the path; represents the risk-oriented pheromone of the edges between the -th iteration and the -th to the -th alternative suppliers; represents the cost-oriented pheromone of the edges between the -th iteration and the -th to the -th alternative suppliers; represents the heuristic factor of the -th alternative supplier; represents the number of alternative suppliers; , , are hyperparameters.

[0014] Preferably, the delivery risk satisfies the expression: ; In the formula, represents the delivery risk of the supplier; represents the on-time delivery rate of the supplier; represents the number of late deliveries, and represents the delay time of the -th late delivery; represents the maximum value function.

[0015] In a second aspect, the present invention provides an optimized video data acquisition system based on artificial intelligence, including a processor and a memory. The memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned supplier management method based on risk assessment is implemented.

[0016] By adopting the above technical solution, a computer program is generated for the above-mentioned supplier management method based on risk assessment and stored in the memory to be loaded and executed by the processor. Thus, a terminal device is manufactured according to the memory and the processor, which is convenient for use.

[0017] The beneficial effects of the present invention are as follows: By constructing a comprehensive evaluation model for the quality risk, delivery risk, financial risk, and compliance risk of suppliers, the present invention realizes the quantitative analysis of the risks of suppliers themselves, thus breaking through the limitations of relying on subjective experience or single-index judgment in traditional procurement decisions and improving the objectivity and accuracy of risk identification. On this basis, a directed graph of risk propagation is established in combination with the upstream and downstream relationships between suppliers, further revealing the conduction path and influence scope of risks in the supply chain network, which helps to identify key nodes and high-risk links and enhances the enterprise's early warning ability for systemic risks.

[0018] The present invention optimizes the procurement path by introducing the ant colony algorithm, combines risk control and cost control, dynamically calculates the risk level, risk propagation intensity, and historical path pheromone distribution of each optional supplier in each round of iteration, and forms a path selection mechanism with "heuristic factor + risk-oriented pheromone + cost-oriented pheromone" as the core, guiding the algorithm to preferentially select supplier combination paths with controllable risks, reasonable costs, and strong stability. This not only improves the overall robustness of the procurement path but also enhances the enterprise's response ability and anti-interference ability in a complex market environment. Description of the Drawings

[0019] Figure 1 It is a flowchart schematically showing a supplier management method based on risk assessment in the present invention. Detailed Embodiments

[0020] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0021] Next, the detailed embodiments of the present invention will be described in conjunction with the accompanying drawings.

[0022] An embodiment of the present invention discloses a supplier management method based on risk assessment. Refer toFigure 1 , including steps S1 - S4: S1. Obtain the basic information of the supplier and historical delivery data.

[0023] Specifically, the basic information includes the supplier name, product type, industry, financial status, upstream suppliers, downstream suppliers, number of violations, etc. Among them, the financial status includes debt ratio, cash flow, current ratio, etc. The number of violations is the number of records of administrative penalties, regulatory notices, legal disputes, etc. received by the supplier.

[0024] The historical delivery data includes product qualification rate, on - time delivery rate, number of delayed deliveries, and the delay time (in days) of each delayed delivery, etc.

[0025] S2. Determine the supplier's own risk indicators based on the basic information of the supplier and historical delivery data.

[0026] Specifically, obtain the quality risk of the supplier according to the product qualification rate of the supplier; obtain the delivery risk of the supplier according to the on - time delivery rate, number of delayed deliveries, and the delay time of each delayed delivery of the supplier; obtain the financial risk of the supplier according to the financial status of the supplier; obtain the compliance risk of the supplier according to the number of violations of the supplier; obtain the supplier's own risk indicators according to the quality risk, delivery risk, financial risk, and compliance risk.

[0027] In one embodiment, the quality risk satisfies the expression: ; In the formula, represents the quality risk of the supplier; represents the product qualification rate of the supplier; when the product qualification rate is higher, the quality risk of the supplier is lower, and vice versa, when the product qualification rate is lower, the quality risk of the supplier is higher.

[0028] In one embodiment, the delivery risk satisfies the expression: ; In the formula, represents the delivery risk of the supplier; represents the on - time delivery rate of the supplier; represents the number of delayed deliveries, represents the th delay time of the delayed delivery; represents the maximum value function, represents selecting and the maximum value of 1. When the number of delayed deliveries is 0, it is stipulated that , so Used to prevent the denominator from being 0.

[0029] In the formula, represents the rate of late delivery. When the on-time delivery rate is lower, the rate of late delivery is higher and the delivery risk is greater; represents the average delay time of delayed delivery. When the average delay time is greater, the impact on delivery is greater. Therefore, in the present invention, is used as the exponent of the rate of late delivery to increase the delivery risk to a certain extent in the form of gamma transformation. When the average delay time is smaller, the impact on delivery is smaller, is closer to 1, making the result of the delivery risk closer to the rate of late delivery . When the average delay time is greater, the impact on delivery is greater, is smaller, making the delivery risk greater based on the rate of late delivery . Therefore, when the on-time delivery rate is smaller and the delay time of each delayed delivery is greater, the delivery risk is greater.

[0030] In one embodiment, the method for obtaining the financial risk is as follows: Set the ideal healthy financial status , where represents the debt ratio in the ideal healthy financial status. When the debt ratio is greater, the debt repayment pressure of the supplier is higher and the financial risk is greater. When the debt ratio is smaller, the capital structure of the supplier is more stable and the risk resistance ability is stronger. Therefore, in this embodiment, is set to 0, indicating that the ideal healthy financial status is debt-free; represents the cash flow in the ideal healthy financial status. When the cash flow is greater, the ability of the supplier to cope with sudden demands or short-term debts is stronger and the risk of capital chain breakage is smaller. When the cash flow is smaller, the supplier may face the risk of falling short of income and operating difficulties. Therefore, in this embodiment, is set to the maximum value of the cash flows of all suppliers in the industry, indicating the level of the supplier with the most abundant funds in the industry; represents the current ratio in the ideal healthy financial status. When the current ratio is greater, the liquidity ability of the supplier is stronger and the short-term debt repayment ability is stronger. When the current ratio is smaller, the liquidity ability of the supplier is poorer and the short-term debt repayment ability is poorer, and there may be a risk of being unable to repay short-term debts. The general healthy current ratio standard in the industry is 2. Therefore, in this embodiment, is set to 2. In other embodiments, the implementer can set the ideal healthy financial status according to the actual implementation situation.

[0031] Furthermore, according to the difference between the financial status of the supplier and the ideal healthy financial status, obtain the financial risk of the supplier: ; In the formula, represents the financial risk of the supplier; represents the debt ratio of the supplier; represents the cash flow of the supplier; represents the current ratio of the supplier; represents the debt ratio under the ideal healthy financial condition, represents the cash flow under the ideal healthy financial condition; represents the current ratio under the ideal healthy financial condition; represents the linear rectifier function, and the expression is .

[0032] In the formula, represents the debt ratio of the supplier and the debt ratio under the ideal healthy financial condition. When the debt ratio of the supplier is larger compared with the debt ratio close to (i.e., close to the debt-free state), it indicates that the debt repayment pressure of the supplier is higher and the financial risk is greater. On the contrary, when represents the cash flow of the supplier and the cash flow under the ideal healthy financial condition. In the formula, is used to normalize the difference so that it falls within the range of [0, 1]. When the cash flow of the supplier is larger compared with the cash flow under the ideal healthy financial condition, it indicates that the cash flow of the supplier is smaller, and the supplier may face the risk of being unable to make ends meet and operating difficulties. At this time, the financial risk is greater; represents the difference between the current ratio of the supplier and the current ratio under the ideal healthy financial condition. When the current ratio of the supplier is smaller compared with the current ratio under the ideal healthy financial condition, it indicates that the liquidity conversion ability of the supplier is poorer and the short-term debt repayment ability is poorer, and there may be a risk of being unable to repay short-term debts. At this time, the financial risk is greater. When the current ratio of the supplier reaches the current ratio under the ideal healthy financial condition, it indicates that the supplier has a strong liquidity conversion ability and short-term debt repayment ability. Therefore, in this embodiment, Perform non - negative truncation processing to prevent the situation where the risk score is lowered when the current ratio is higher than the ideal value. In the formula, use to perform normalization so that it falls within the range of [0, 1].

[0033] In one embodiment, the compliance risk satisfies the expression: ; wherein, represents the compliance risk of the supplier; represents the number of violations of the supplier, that is, the number of records such as administrative penalties, regulatory notifications, legal disputes, etc. received by the supplier; represents the hyperbolic tangent function, which is used to normalize the number of violations . When the number of violations of the supplier is more, its historical behavior indicates that its compliance awareness is weaker, the possibility of future violations is higher, and the compliance risk of the supplier is greater. The greater the compliance risk of the supplier.

[0034] In one embodiment, the self - risk index satisfies the expression: ; In the formula, represents the self - risk index of the supplier; represents the quality risk of the supplier; represents the delivery risk of the supplier; represents the financial risk of the supplier; represents the compliance risk of the supplier; represents the normalization function. In this embodiment, linear normalization is adopted, and the financial risk of the current supplier is linearly normalized according to the financial risks of all suppliers. In other embodiments, the implementer can select the normalization method according to the actual implementation situation; and and and respectively represent the weights of the quality risk, delivery risk, financial risk, and compliance risk. In this embodiment, the attention degrees to the quality risk, delivery risk, financial risk, and compliance risk are the same. Therefore . In other embodiments, the implementer can adjust the weights of the quality risk, delivery risk, financial risk, and compliance risk according to industry characteristics. For example, the medical industry can increase the weight of the quality risk, and the electronic manufacturing industry can increase the weight of the delivery risk. When the quality risk, delivery risk, financial risk, and compliance risk of the supplier are greater, the supplier is more likely to have key risk events such as supply interruption, quality problems, capital chain breakage, legal disputes, etc. At this time, the self - risk index R of the supplier is greater, indicating that the overall operational stability of the supplier is worse.

[0035] S3. Construct a directed graph of risk propagation based on the upstream and downstream relationships between suppliers, and obtain the risk propagation index between any two suppliers according to the directed graph of risk propagation.

[0036] Specifically, constructing a directed graph of risk propagation based on the upstream and downstream relationships between suppliers includes: Regarding each supplier as a node respectively, for any two suppliers and , in response to supplier being the upstream supplier of supplier (i.e., supplier provides raw materials, components, etc. to supplier ), establish a directed edge pointing from the node corresponding to supplier to the node corresponding to supplier , and determine the weight of the directed edge pointing from supplier to supplier according to the product proportion of supplier supplying to supplier . pointing to supplier .

[0037] By judging the upstream and downstream relationships between any two suppliers, obtain all the directed edges, and form a directed graph based on the nodes, directed edges, and the weights of the directed edges, denoted as the directed graph of risk propagation.

[0038] Among them, the weight of the directed edge satisfies the expression: ; Among them, represents the weight of the directed edge pointing from supplier to supplier ; represents the quantity of the th product supplied by supplier to supplier ; represents the total quantity of the th product supplied by all upstream suppliers of supplier to supplier ; represents the number of product types supplied by supplier to supplier ; represents the maximum value function.

[0039] When is close to , it means that supplier is the main supplier of the th product. When is much smaller than , it means that the supplier Supplier is the secondary supplier of the th product. When the supplier is the main supplier of a certain product of the supplier , is close to 1, and the greater the weight of the directed edge, the stronger the supply dependence of the supplier on the supplier . Once the supplier has a supply interruption, quality, or delivery problem, the impact on the supplier is more serious; if among all the product categories supplied by the supplier to the supplier , the supplier is only the secondary supplier of the supplier , is close to 0, and the smaller the weight of the directed edge, the weaker the impact of the supplier on the supplier . Even if the supplier has a problem, the supplier is easier to be replaced through other channels, and the risk propagation ability is lower.

[0040] Furthermore, according to the risk propagation directed graph, obtain the risk propagation index between any two suppliers, including: For any two suppliers and , obtain all the paths from the supplier to the supplier in the risk propagation directed graph, and take each of them as a risk propagation path. According to 's own risk index and the weights of all the directed edges in each risk propagation path, obtain the risk propagation index from the supplier to the supplier : ; In the formula, represents the risk propagation index from the supplier to the supplier ; represents the own risk index of the supplier ; represents the th path from the supplier to the supplier The weight of the th directed edge in the th risk propagation path; Indicates the number of directed edges in the th risk propagation path from supplier to supplier in the risk propagation digraph; Indicates the number of risk propagation paths from supplier to supplier in the risk propagation digraph; Indicates the risk propagation attenuation coefficient of the th risk propagation path from supplier to supplier in the risk propagation digraph; Indicates the risk propagation intensity of the th risk propagation path from supplier to supplier in the risk propagation digraph; When the risk propagation intensity of each risk propagation path from supplier to supplier is greater, the risk propagation index from supplier to supplier is greater. On the contrary, when the risk propagation intensity of each risk propagation path from supplier to supplier is smaller, the risk propagation index from supplier to supplier

[0041] It should be noted that when there is no risk propagation path from supplier to supplier in the risk propagation digraph, it is stipulated that the risk propagation index from supplier to supplier is 0.

[0042] S4. Construct a procurement path using the ant colony algorithm based on the self-risk index of the supplier and the risk propagation index between suppliers.

[0043] Each supplier is regarded as a node, and the ant colony algorithm is used to construct the procurement path. Among them, in each round of iteration of the ant colony algorithm, all suppliers that have not been added to the current path are regarded as optional suppliers respectively. According to the own risk index of the optional suppliers and the risk propagation index from the suppliers corresponding to the nodes in the current path to the optional suppliers, the heuristic factor of the optional suppliers is determined; according to the sum of the own risk coefficients of the suppliers included in each path obtained in the previous round of iteration, the risk-oriented pheromone of the edge between the supplier corresponding to the current node in the current path and the optional suppliers is determined; according to the sum of the unit procurement costs of the suppliers included in each path obtained in the previous round of iteration, the cost-oriented pheromone of the edge between the supplier corresponding to the current node in the current path and the optional suppliers is determined; according to the risk-oriented pheromone, the cost-oriented pheromone and the heuristic factor, the selection probability of the optional suppliers is determined; according to the selection probability, a supplier is selected from all the optional suppliers by the method of unequal probability sampling as the next node of the current path. Repeat this path expansion process until the types and quantities of products that the suppliers in the current path can provide meet the current procurement requirements, or when there are no unvisited optional suppliers, stop the construction of the current path.

[0044] The optimal path obtained after several rounds of iteration is used as the procurement path, and structured procurement decisions and order allocations are made according to the set of suppliers included in the procurement path.

[0045] It should be noted that the total amount of pheromone in the ant colony algorithm of the present invention is set to 1, the pheromone evaporation coefficient is set to 0.5, and the maximum number of iterations is set to 100. Implementers can also set according to the actual implementation situation.

[0046] In one embodiment, the heuristic factor satisfies the expression: ; where represents the heuristic factor of the th optional supplier, represents the own risk index of the th optional supplier; represents the risk propagation index from the supplier corresponding to the th node in the current path to the th optional supplier; represents the number of existing nodes in the current path; represents the exponential function with the natural constant as the base, which is used to perform a negative correlation mapping on .

[0047] When the own risk index of the th optional supplier The larger it is, the higher the probability that risk events such as quality problems, delivery delays, or supply disruptions occur for the supplier. At this time, its heuristic factor is smaller, thus encouraging the ant to select a supplier with lower risk and more stable performance; if the risk propagation index from the suppliers corresponding to the nodes in the current path to the th optional supplier is larger, it indicates that if a risk event such as a quality problem, delivery delay, or supply disruption occurs for a certain supplier in the current path, it may have a negative impact on the th optional supplier through the risk propagation mechanism, that is, the th supplier has a higher associated risk. At this time, the heuristic factor of the th optional supplier is smaller, thus encouraging the ant to select a supplier with less influence from the existing nodes in the path, higher independence, and stronger anti-interference ability.

[0048] In one embodiment, the risk-oriented pheromone satisfies the expression: ; Denote the current iteration as the th iteration, and denote the supplier corresponding to the current node in the current path as supplier . Then in the formula, represents the risk-oriented pheromone of the edge between supplier and the th to the th optional suppliers after the th iteration; represents the risk-oriented pheromone of the edge between supplier and the th to the th optional suppliers after the th iteration; represents the increment of the risk-oriented pheromone of the edge between supplier and the th optional suppliers after the

[0049] Among them, the risk-oriented information increment satisfies the expression: ; In the formula, represents the increment of the risk-oriented pheromone of the edge between supplier and the th to the th optional suppliers after the th iteration; represents the result obtained after the th iteration, including from supplier The number of paths between edges of represents the total amount of pheromone; In the result obtained in the th round of iteration, it includes the suppliers corresponding to all nodes in the th path between the th optional supplier and the th path, which is the sum of the self-risk coefficients of the suppliers corresponding to all nodes; represents the risk-oriented pheromone increment brought by the th path. When the sum of the self-risk coefficients of the suppliers corresponding to all nodes in the th path is smaller, the risk-oriented pheromone increment brought by the th path is larger. When the risk-oriented pheromone increment brought by all paths between the th supplier and the th optional supplier is larger, the risk-oriented pheromone increment between the th supplier and the th optional supplier is larger, thus encouraging the ants to preferentially select paths with lower risks and stronger node stability.

[0050] It should be noted that in this embodiment, the initial risk-oriented pheromone between any two suppliers is 1. In other embodiments, the implementer can set the initial risk-oriented pheromone according to the actual implementation situation.

[0051] In one embodiment, the cost-oriented pheromone satisfies the expression: ; Denote the current iteration as the th round of iteration, and denote the supplier corresponding to the current node in the current path as supplier . Then in the formula, represents the cost-oriented pheromone between the th round of iteration and the th to the th optional suppliers; represents the cost-oriented pheromone between the th round of iteration and the th to the th optional suppliers; represents the cost-oriented pheromone increment between the th round of iteration and the th to the th optional suppliers; represents the pheromone evaporation coefficient.

[0052] Among them, the cost-oriented information increment satisfies the expression: ; In the formula, represents the cost-oriented pheromone increment of the edge between the -th and the -th to -th alternative suppliers after the -th iteration; represents the number of paths including the edge between the -th and the -th alternative suppliers in the result obtained after the -th iteration, represents the total amount of pheromone; represents the sum of the unit procurement costs of the suppliers corresponding to all nodes in the -th to -th path including the edge between the -th and the -th alternative suppliers in the result obtained after the -th iteration; represents the cost-oriented pheromone increment brought by the -th path. When the sum of the unit procurement costs of the suppliers corresponding to all nodes in the -th path is smaller, the cost-oriented pheromone increment brought by the -th path is larger. When the cost-oriented pheromone increments brought by all paths including the edge between the -th and the -th alternative suppliers are larger, the cost-oriented pheromone increment of the edge between the

[0053] It should be noted that in this embodiment, the initial cost-oriented pheromone of the edge between any two suppliers is 1. In other embodiments, the implementer can set the initial cost-oriented pheromone according to the actual implementation situation.

[0054] In one embodiment, the selection probability satisfies the expression: ; Denote the current iteration as the -th iteration, and denote the supplier corresponding to the current node in the current path as supplier . Then in the formula, represents the selection probability of the -th alternative supplier as the next node of the path; represents the situation after the -th iteration, and the supplier to the risk-oriented pheromone on the edges between the alternative suppliers; indicating after the th round of iteration, the cost-oriented pheromone on the edges between the to the alternative suppliers; indicating the heuristic factor of the th alternative supplier; indicating the number of alternative suppliers; , , are hyperparameters, used to control the relative importance of the risk-oriented pheromone, used to control the relative importance of the cost-oriented pheromone, used to control the relative importance of the heuristic factor. In this embodiment, the relative importance of the risk-oriented pheromone, the cost-oriented pheromone, and the heuristic factor is the same. Therefore, set . In other embodiments, the implementer can set the hyperparameters , , according to the actual implementation situation.

[0055] An embodiment of the present invention also discloses an artificial intelligence-based video data optimized acquisition system, including a processor and a memory. The memory stores computer program instructions, which, when executed by the processor, implement a risk assessment-based supplier management method according to the present invention.

[0056] The above system further includes other components well-known to those skilled in the art, such as a communication bus and a communication interface. Their settings and functions are known in the art, so they will not be described in detail here.

Claims

1. A supplier management method based on risk assessment, characterized in that, Including: Obtain the own risk indicators of suppliers according to the quality risk, delivery risk, financial risk and compliance risk of the suppliers; Construct a directed risk propagation graph according to the upstream and downstream relationships between suppliers, and obtain the risk propagation index between any two suppliers according to the directed risk propagation graph; Use the ant colony algorithm to construct a procurement path. In each iteration of the ant colony algorithm, determine the heuristic factor of the optional supplier according to the own risk index of the optional supplier and the risk propagation index from the supplier corresponding to each node in the current path to the optional supplier; Determine the risk-oriented pheromone of the edge between the supplier corresponding to the current node in the current path and the optional supplier according to the sum of the own risk coefficients of the suppliers included in each path obtained in the previous iteration; determine the cost-oriented pheromone of the edge between the supplier corresponding to the current node in the current path and the optional supplier according to the sum of the unit procurement costs of the suppliers included in each path obtained in the previous iteration; determine the selection probability of the optional supplier according to the risk-oriented pheromone, cost-oriented pheromone and heuristic factor; select a supplier from all optional suppliers as the next node of the current path according to the selection probability.

2. The supplier management method based on risk assessment according to claim 1, wherein The method for obtaining the own risk indicators of the suppliers is as follows: Obtain the quality risk of the supplier according to the product qualification rate of the supplier; obtain the delivery risk of the supplier according to the on-time delivery rate of the supplier, the number of delayed deliveries and the delay time of each delayed delivery; obtain the financial risk of the supplier according to the financial status of the supplier; obtain the compliance risk of the supplier according to the number of violations of the supplier; Perform weighted summation on the quality risk, delivery risk, financial risk and compliance risk to obtain the own risk indicators of the suppliers.

3. The supplier management method based on risk assessment according to claim 1, wherein The construction of the directed risk propagation graph includes: Regarding each supplier as a node respectively, for any two suppliers and , in response to supplier being the upstream supplier of supplier , establish a directed edge from supplier to supplier . Determine the weight of the directed edge from supplier to supplier to supplier : : , represents the quantity of the -th product supplied by supplier to supplier ; represents the total quantity of the -th product supplied by all upstream suppliers of supplier to supplier ; represents the number of product types supplied by supplier to supplier ; represents the maximum value function. Form a directed graph based on all nodes, directed edges, and the weights of the directed edges, denoted as the risk propagation directed graph.

4. The supplier management method based on risk assessment according to claim 1, characterized in that, The method for obtaining the risk propagation index is as follows: For any two suppliers and , obtain all paths from supplier to supplier in the risk propagation directed graph, and respectively regard each of them as a risk propagation path; Supplier To the supplier The risk propagation index Satisfies the expression: , Indicates the self - risk index of the supplier ; Indicates the weight of the th directed edge in the th risk propagation path from the supplier to the supplier ; Indicates the number of directed edges in the th risk propagation path from the supplier to the supplier ; Indicates the number of risk propagation paths from the supplier to the supplier .

5. The supplier management method based on risk assessment according to claim 1, characterized in that, The heuristic factor satisfies the expression: ； Among them, represents the heuristic factor of the th optional supplier, represents the self-risk index of the th optional supplier; represents the risk propagation index from the supplier corresponding to the th node in the current path to the th optional supplier; represents the number of existing nodes in the current path; represents the exponential function with the natural constant as the base.

6. The supplier management method based on risk assessment according to claim 1, characterized in that The risk-oriented pheromone satisfies the expression: ； Denote the current iteration as the th iteration, and denote the supplier corresponding to the current node in the current path as Supplier . Wherein represents the risk-oriented pheromone of the edges between Supplier and Supplier to Supplier after the th iteration; represents the risk-oriented pheromone of the edges between Supplier and Supplier to Supplier after the th iteration; represents the pheromone evaporation coefficient; represents the increment of the risk-oriented pheromone of the edges between Supplier and Supplier after the th iteration: , where represents the number of paths of the edges between Supplier and Supplier in the result obtained after the th iteration, represents the total pheromone; represents the sum of the self-risk coefficients of the suppliers corresponding to all nodes in the th path of the edges between Supplier and Supplier in the result obtained after the th iteration.

7. A supplier management method based on risk assessment according to claim 1, characterized in that, The cost-oriented pheromone satisfies the expression: ； Denote the current iteration as the -th iteration, and denote the supplier corresponding to the current node in the current path as Supplier . Wherein represents the cost-oriented pheromone of the edge between Supplier and Supplier after the -th iteration; represents the cost-oriented pheromone of the edge between Supplier and Supplier after the -th iteration; represents the pheromone evaporation coefficient; represents the increment of the cost-oriented pheromone of the edge between Supplier and Supplier after the -th iteration: , represents the number of paths of the edge between Supplier and Supplier to Supplier in the result obtained in the -th iteration, represents the total pheromone; represents the sum of the unit procurement costs of the suppliers corresponding to all nodes in the -th path of the edge between Supplier and Supplier in the result obtained in the 8. A supplier management method based on risk assessment according to claim 1, characterized in that The selection probability of the optional supplier satisfies the expression: ； Denote the current iteration as the -th iteration, and denote the supplier corresponding to the current node in the current path as Supplier . In the formula, represents the selection probability of the -th alternative supplier as the next node of the path; represents the risk-oriented pheromone of the edge between Supplier and the -th to -th alternative suppliers after the -th iteration; represents the cost-oriented pheromone of the edge between Supplier and the -th to -th alternative suppliers after the -th iteration; represents the heuristic factor of the -th alternative supplier; represents the number of alternative suppliers; are hyperparameters.

9. The supplier management method based on risk assessment according to claim 2, wherein The delivery risk satisfies the expression: ； Wherein, represents the delivery risk of the supplier; represents the on-time delivery rate of the supplier; represents the number of late deliveries, represents the delay time of the represents the maximum value function.

10. An artificial intelligence-based video data optimized acquisition system, characterized in that, Including: A processor and a memory, where the memory stores computer program instructions, and when the computer program instructions are executed by the processor, a supplier management method based on risk assessment according to any one of claims 1-9 is implemented.

Images