Supply Chain Resource Optimization Method and System Based on Atlas Analysis

Abstract

The present invention discloses a supply chain resource optimization method and system based on graph analysis, including the following steps: constructing a supply chain graph and updating the supply chain graph based on real-time supply chain data; identifying critical paths, inefficient paths and confirming bottleneck nodes in the supply chain graph through graph analysis algorithms; generating an initial optimization strategy set based on the critical paths and bottleneck nodes; receiving user interaction signals, determining the type of optimization operation according to the user interaction signals, and dynamically correcting the initial optimization strategy set based on the type of optimization operation; executing the initial optimization strategy set after dynamic correction, and updating the supply chain graph based on the execution result feedback. The present invention has the following advantages and effects: realizing dynamic global analysis and real-time response of the supply chain network, improving the efficiency of resource allocation and reducing the risk of logistics interruption in a complex supply chain environment.

Description

Technical Field

[0001] The present invention relates to the field of supply chain resource optimization, and particularly to a supply chain resource optimization method and system based on graph analysis. Background Art

[0002] With the development of globalization and informatization, the complexity and dynamics of the supply chain have increased significantly. Especially in large-scale and cross-regional supply chains, managers face many challenges such as multi-level supply chain nodes, resource flows, and demand fluctuations. Effective supply chain management is crucial for enhancing the competitiveness of enterprises, reducing operating costs, and improving customer satisfaction. To address these challenges, more and more enterprises are beginning to optimize the allocation of supply chain resources by means of advanced technologies such as information technology, artificial intelligence, big data, and cloud computing. In this context, traditional supply chain optimization methods have gradually revealed key technical defects: it is difficult to achieve dynamic global analysis and real-time response of complex supply chain networks, resulting in low resource allocation efficiency and lagging risk response. Existing technologies mostly make decisions based on static rules or local data, such as single-node inventory levels and fixed transportation routes, lacking the ability to comprehensively analyze the dynamic correlation relationships between multi-level resource nodes, real-time supply and demand changes, and external environmental disturbances. In the face of sudden order surges or logistics interruptions, traditional systems cannot quickly identify key path bottlenecks, nor can they achieve optimal allocation of resources through topological structure reorganization and dynamic weight adjustment, resulting in lagging response, ineffective local optimization, and insufficient overall supply chain resilience. This defect seriously restricts the resource utilization efficiency and risk resistance ability of enterprises in complex business environments. Summary of the Invention

[0003] The purpose of the present invention is to provide a supply chain resource optimization method and system based on graph analysis to solve the problems raised in the background art.

[0004] The above technical objective of the present invention is achieved through the following technical solutions:

[0005] The present invention provides a supply chain resource optimization method based on graph analysis, including the following steps:

[0006] S100. Construct a supply chain graph and update the supply chain graph based on real-time supply chain data; wherein, the supply chain graph includes multiple resource nodes, the association relationships between the resource nodes, and the attribute information of each resource node, and the real-time supply chain data includes order information, inventory status, logistics dynamics, and external environment data;

[0007] S200. Identify key paths, inefficient paths, and confirm bottleneck nodes in the supply chain graph through graph analysis algorithms;

[0008] S300. Generate an initial optimization strategy set based on the critical path and bottleneck nodes; the initial optimization strategy set includes a resource reallocation plan, a logistics path adjustment plan, and a priority adjustment plan;

[0009] S400. Receive user interaction signals, determine the type of optimization operation according to the user interaction signals, and dynamically correct the initial optimization strategy set based on the type of optimization operation;

[0010] S500. Execute the initial optimization strategy set after dynamic correction, and update the supply chain map based on the execution result feedback.

[0011] By adopting the above technical solutions, through the acquisition of real-time supply chain data and map analysis, it can help enterprises dynamically identify critical paths, inefficient paths, and bottleneck nodes; traditional supply chain optimization methods usually rely on static rules or local data for decision-making, and it is difficult to cope with real-time changes and complex environmental factors. However, this application enables the supply chain system to perceive external changes in real time and make rapid responses by continuously updating the supply chain map. First, by constructing a supply chain map, each resource node and its attribute information in the supply chain can be presented one by one. This information-based map can provide comprehensive supply chain transparency and better understand the status of each link and its mutual relationship. Second, the application of map analysis algorithms helps to discover critical paths and inefficient paths and further optimize these paths. The dynamic correction of the optimization plan and the user interaction function further enhance the flexibility and adaptability of this method. By collecting user interaction signals in real time and adjusting the type of optimization operation according to the feedback, this method can flexibly handle different supply chain problems; users can select different optimization operations according to the actual situation, so as to ensure that the supply chain system can respond quickly and continuously optimize.

[0012] A further setting is that the specific implementation of constructing the supply chain map in S100 includes:

[0013] S110. Collect historical supply chain data, where the historical supply chain data includes supplier information, distribution channels, warehousing nodes, and customer demand records;

[0014] S120. Extract multi-dimensional features from the historical supply chain data to generate attribute information of resource nodes; the attribute information includes resource type, supply capacity, demand priority, resource utilization rate, historical supply-demand matching rate, and dynamic weight; where:

[0015] The demand priority is dynamically assigned according to the order urgency, customer level, and contract constraint conditions to obtain the corresponding demand priority coefficient;

[0016] The dynamic weight is composed of the weighted sum of three parts, specifically including:

[0017] Divide the historical supply - demand matching rate of the resource node by the preset maximum matching rate benchmark value, and then multiply by the weight adjustment factor α to obtain the proportion of the historical supply - demand matching rate, which serves as the first part;

[0018] Divide the current demand priority coefficient by the average priority coefficient, and then multiply by the weight adjustment factor β to obtain the proportion of the current demand priority, which serves as the second part;

[0019] Divide the resource utilization rate by the total resource capacity, and then multiply by the weight adjustment factor γ to obtain the proportion of the resource utilization rate, which serves as the third part;

[0020] Add the first part, the second part, and the third part together to obtain the dynamic weight;

[0021] S130. Construct the topological structure of the supply chain map based on the graph database; in the topological structure:

[0022] Resource nodes are connected by directed edges, and the directed edges represent the resource flow direction and the association strength;

[0023] Bind the attribute information to each resource node to form a complete node description including resource type tags, supply capacity values, demand priority coefficients, resource utilization rates, historical supply - demand matching rates, and dynamic weight parameters.

[0024] By adopting the above - mentioned technical solutions, by collecting historical supply chain data and performing multi - dimensional feature extraction on the attributes of resource nodes based on these data, accurate attribute descriptions can be provided for each resource node; in terms of the calculation of the dynamic weight, this method combines multiple factors such as the historical supply - demand matching rate, demand priority, and resource utilization rate, and generates the dynamic weight through weighted summation. This kind of weight considers multiple aspects of factors, enabling each resource node in the map to be dynamically adjusted according to its current state and importance, providing a more accurate data basis for subsequent map analysis and path optimization; through the construction of the topological structure of the graph database, the relationships between resource nodes are expressed more clearly; the nodes are connected by directed edges, and the directed edges not only represent the direction of resource flow but also can reflect the association strength between resources; each node binds its attribute information to form a complete resource description, enabling the map to more accurately represent the state and behavior of each resource in the supply chain. The construction of this topological structure not only improves the readability of the map but also enhances the system's management ability for complex supply chains, helping managers make better decisions; ultimately, it provides important support for the real - time monitoring, path optimization, and bottleneck node identification of the supply chain, enabling the optimization process to make rapid and accurate decisions based on real - time data, and significantly improving the operation efficiency of the supply chain.

[0025] A further setting is that the association strength in S130 is specifically as follows:

[0026] S130.1. Based on the historical supply chain data collected in S110, count the resource interaction frequency and interaction volume between adjacent resource nodes;

[0027] S130.2. Based on the resource interaction frequency, resource interaction volume, and node attribute information generated in S120, calculate the association strength value; the calculation process of the association strength value is as follows;

[0028] Divide the resource interaction frequency by the set maximum interaction frequency reference value to obtain the normalized interaction frequency ratio;

[0029] Divide the resource interaction volume by the set maximum interaction volume reference value to obtain the normalized interaction volume ratio;

[0030] Take the average value of the supply capabilities of the source node and the target node as the contribution value of the supply capability to the association strength;

[0031] Multiply the above three parts of the results by the preset weight coefficients respectively, and satisfy the condition that the sum of the weight coefficients is 1. The association strength value is the sum of the three parts;

[0032] S130.3. Assign weights to the directed edges according to the association strength value:

[0033] If the association strength value is greater than or equal to the first threshold, assign a high association strength label to the directed edge;

[0034] If the association strength value is between the second threshold and the first threshold, assign a medium association strength label to the directed edge;

[0035] If the association strength value is less than the second threshold, assign a low association strength label to the directed edge;

[0036] S130.4. Mark the critical links based on the dynamic weight adjustment threshold in S140:

[0037] According to the dynamic weights of the resource nodes calculated in S140, readjust the first threshold so that the first threshold increases as the importance of the nodes increases;

[0038] Only when the association strength value reaches or exceeds the adjusted first threshold, mark the association relationship as a critical link.

[0039] By adopting the above technical solutions, through the statistics of historical supply chain data and the calculation of the resource interaction frequency and interaction volume between adjacent resource nodes, the association strength values between nodes can be obtained. These association strength values can not only reflect the degree of closeness of the relationship between nodes, but also provide a scientific basis for optimizing path selection and resource allocation; by comprehensively considering the resource interaction frequency, interaction volume and node attribute information, the association strength value of each directed edge is calculated. This calculation method takes into account multiple factors such as the resource flow situation, resource utilization rate and supply capacity between nodes, and can more accurately evaluate the association strength between nodes; through the analysis of the association strength values, the critical links can be identified, and optimization decisions can be made based on these links to ensure the stability and smoothness of the supply chain; in addition, by dynamically adjusting the threshold to mark the critical links, the marking threshold can be dynamically adjusted according to the importance of the nodes, so as to ensure that the resource supply of the critical links is preferentially guaranteed. This dynamic adjustment mechanism not only improves the accuracy of optimization, but also can cope with the possible changes in the supply chain.

[0040] A further setting is that the S200 specifically includes the following sub-steps:

[0041] S210. Traverse the supply chain map using the depth-first search algorithm to extract all complete paths from the supplier to the end customer;

[0042] S220. Calculate the resource flow efficiency of each path;

[0043] S230. Mark the critical paths and inefficient paths based on the resource flow efficiency; specifically including the following sub-steps:

[0044] Compare the resource flow efficiency value with the preset efficiency threshold range, and the preset efficiency threshold range includes a high-efficiency threshold and an inefficient threshold;

[0045] If the resource flow efficiency value is greater than or equal to the high-efficiency threshold, mark the corresponding path as a critical path;

[0046] If the resource flow efficiency value is less than the inefficient threshold, mark the corresponding path as an inefficient path;

[0047] S240. Perform differential analysis on the critical paths and inefficient paths and confirm the bottleneck nodes; specifically including the following sub-steps:

[0048] Extract the associated edge with the highest association strength value in the critical path, mark it as the core link, and allocate additional resource redundancy to the core link to ensure its stability;

[0049] For each inefficient path, extract the dynamic weights of all nodes in the path and filter out the node with the lowest dynamic weight; if the dynamic weights of multiple nodes are the same and are all the lowest values, further compare their historical supply-demand matching rates, and confirm the node with the lowest matching rate as the bottleneck node;

[0050] Associate and store the core link information of the critical path and the bottleneck node information of the inefficient path.

[0051] By adopting the above technical solution, the application of the depth-first search algorithm can comprehensively traverse the supply chain map and extract all complete paths from suppliers to end customers. This process can not only comprehensively analyze the global situation of resource flow but also deeply explore potential problems in each path. Through this comprehensive path traversal, the performance of each path can be evaluated in detail, and bottlenecks or resource waste problems that may exist can be discovered in a timely manner; based on the calculated resource flow efficiency, by comparing it with the preset efficiency threshold range, mark the critical path and the inefficient path, providing a clear direction for subsequent optimization work; the implementation of the alienation analysis ensures that different optimization measures can be taken for different types of paths. The optimization of the critical path usually focuses on ensuring its stability and smoothness, while the optimization of the inefficient path focuses more on reducing resource waste and improving resource flow efficiency, ensuring that resources can reach the end node more quickly. In this process, by identifying bottleneck nodes and taking corresponding resource adjustment measures, the resource flow efficiency of the path can be effectively improved.

[0052] A further setting is that the calculation of the resource flow efficiency in S220 includes the following sub-steps:

[0053] Extract the dynamic weights of all nodes in the path and calculate the geometric mean of the node dynamic weights;

[0054] Extract the weight labels of all associated edges in the path and calculate the arithmetic mean of the association strength values;

[0055] Statistical path physical length, where the physical length is defined as the sum of the geographical distances between adjacent nodes in the path;

[0056] Calculate the resource flow efficiency through the following comprehensive evaluation logic:

[0057] Multiply the geometric mean of the node dynamic weights by the arithmetic mean of the association strength values to obtain a comprehensive evaluation value;

[0058] Add the path physical length to a preset anti-zero constant, where the anti-zero constant is a very small positive number used to avoid calculation anomalies;

[0059] Divide the comprehensive evaluation value by the length adjustment value to obtain the resource flow efficiency value; the higher the resource flow efficiency value, the better the resource transfer ability and stability of the path.

[0060] By adopting the above technical solutions, it is possible to more comprehensively and accurately evaluate the efficiency of each path and provide strong data support for subsequent optimization decisions; the geometric mean of the dynamic weights of nodes can reflect the matching situation of resource supply and demand of each node in the path, and nodes with stronger supply capabilities will have a positive impact on the overall efficiency of the path; the resource flow between nodes with higher association strength is usually smoother, so the overall efficiency of the path is also higher; on the contrary, if the association strength between nodes in the path is low, there may be problems such as unsmooth resource flow or lag in information transmission, resulting in a decrease in path efficiency; long paths may lead to an increase in transportation time, thereby affecting the timely delivery of resources. Therefore, when calculating efficiency, the geographical distance of the path needs to be taken into account; the comprehensive evaluation logic obtains the comprehensive evaluation value of each path by weighted summing the above multiple factors. Subsequently, the physical length of the path will be added to a preset anti-zero constant to avoid abnormal situations in the calculation. Finally, the comprehensive evaluation value is divided by the length adjustment value to obtain the resource flow efficiency value; the higher this efficiency value, the stronger the resource transfer ability and the higher the stability of the path; on the contrary, it indicates that there is room for optimization of this path; this comprehensive evaluation model can not only accurately calculate the resource flow efficiency of each path, but also provide strong support for subsequent optimization decisions.

[0061] A further setting is that the specific implementation of generating the initial optimization strategy set in S300 includes:

[0062] S310. Generate a resource reallocation plan for inefficient paths; specifically including the following sub-steps:

[0063] Extract the inefficient paths marked in S230 and S240 and their corresponding bottleneck node information;

[0064] Traverse the adjacent nodes of the bottleneck nodes and screen out the adjacent nodes whose redundant resources meet the preset conditions; the screening process is specifically that the current resource utilization rate of the adjacent node is lower than the preset utilization rate threshold, and its supply capacity is greater than the current load;

[0065] Calculate the resource gap of the bottleneck node, that is, the difference between the current demand and the current supply;

[0066] Based on the redundant resource capacity and association strength value of the adjacent node, allocate resource requirements according to the following rules:

[0067] Give priority to selecting adjacent nodes with high association strength values;

[0068] The allocation amount does not exceed the set percentage threshold of the redundant resources of the adjacent node;

[0069] Generate a resource transfer instruction and add the resource transfer instruction to the initial optimization strategy set;

[0070] S320. Generate a logistics path adjustment plan for the critical link; specifically including the following sub-steps:

[0071] Extract the critical path marked in S230 and its core link information;

[0072] Obtain the physical length of the current logistics path of the core link and the external environment data;

[0073] If the physical length of the current logistics path exceeds the preset length threshold or the external environment risk level is higher than the preset risk threshold, perform the following adjustments:

[0074] Execute the path shortening strategy, specifically:

[0075] Through the topological structure of S130, search for links in the alternative path whose association strength value is not lower than the original path and whose physical length is shorter; if no alternative path can be found, retain the original path, but trigger a warning signal and generate a path optimization suggestion report, and the optimization suggestion report includes increasing logistics capacity, temporarily enabling standby nodes or adjusting the transportation time period;

[0076] Execute the risk avoidance strategy, specifically:

[0077] Based on the supplier information collected in S110 and the number of historical accidents and total transportation times recorded in the warehouse node data, calculate the risk coefficient of the nodes in the path, and calculate the risk coefficient of the nodes in the path;

[0078] Screen nodes with a risk coefficient lower than the preset threshold and adjacent in geographical location as standby nodes;

[0079] Replace the high-risk nodes with standby nodes, and recalculate the association strength value and physical length of the path after replacement;

[0080] Generate an optimized logistics path instruction, and add the logistics path instruction to the initial optimization strategy set.

[0081] By adopting the above technical solutions, different optimization means are taken for different problems, so as to achieve the precise optimization of the supply chain; for inefficient paths, a resource reallocation plan is generated to reduce resource waste and improve path efficiency; by analyzing the bottleneck nodes in the inefficient paths, adjacent nodes with redundant resources meeting the preset conditions can be screened out, and the bottleneck problem can be alleviated by adjusting the resource allocation. In this process, adjacent nodes with a higher correlation strength can be preferentially selected for resource allocation to ensure that resources can flow to the places where they are most needed, thereby reducing delays and waste on the path; for critical paths, a logistics path adjustment plan is generated. When optimizing the path, the physical length of the current logistics path and the risk factors of the external environment will be considered. If the transportation time of the current path is relatively long, or the risk level of the external environment is relatively high, alternative paths will be searched according to the topological structure, and paths with a higher correlation strength value and a shorter physical length will be preferentially selected; if no suitable alternative path is found, a warning signal will be sent to remind the manager to take corresponding risk avoidance measures, such as increasing transportation capacity, enabling standby nodes or adjusting the transportation time period; in terms of priority adjustment, the priority of resources will be dynamically adjusted according to the urgency of the order, the customer level and the contract constraint conditions. This adjustment ensures that high-priority requirements can be met in a timely manner, thereby improving customer satisfaction and resource utilization rate; for the situation of priority conflicts, the allocation order of resources can also be dynamically adjusted according to different priority rules to ensure that resources can most effectively serve the most urgent needs.

[0082] A further setting is that in S120, the demand priority is dynamically assigned according to the order urgency, customer level and contract constraint conditions, and the specific process of obtaining the corresponding demand priority coefficient is as follows:

[0083] The orders are divided into four levels: urgent, high, medium and low, and the weight of the order urgency is quantified based on the contract constraint conditions;

[0084] The customers are divided into two categories: VIP and ordinary, and the customer level weights are assigned according to the historical cooperation duration, order scale and performance record;

[0085] The demand priority coefficient is calculated through the following priority coefficient generation rules:

[0086] Basic weights are assigned to the order urgency and customer level respectively;

[0087] The basic weights are weighted and summed with the contract liquidated damages ratio to generate the priority coefficient;

[0088] The priority coefficient is linked and adjusted with the node dynamic weight to ensure that high-priority resource nodes are preferentially matched with high-priority requirements;

[0089] The specific implementation of generating the initial optimization strategy set in S300 further includes:

[0090] S330. Generate a priority adjustment plan for demand priority conflicts; specifically including the following sub-steps:

[0091] Receive the updated order information in real time from S110, and analyze the order urgency and customer level;

[0092] Based on the priority coefficient generation rule defined in S120, recalculate the demand priority coefficient of the conflict node;

[0093] If there are demand conflicts for multiple orders on the same resource node, handle them according to the following rules:

[0094] Give priority to ensuring the demands with a higher classification of order urgency;

[0095] If the urgency classifications are the same, determine the priority according to the customer level weights;

[0096] If the urgency classifications and customer levels are both the same, sort them in descending order according to the contract liquidated damages ratio;

[0097] Then, according to the updated priority coefficient, dynamically adjust the sorting of the demand queue of the resource node:

[0098] Pin the order with the highest priority coefficient to the top and allocate reserved resource capacity for it;

[0099] Generate a delay processing suggestion for the order with the second highest priority coefficient, including the estimated processing time and recommended alternative resource nodes;

[0100] Finally, add the adjusted priority instruction to the initial optimization strategy set.

[0101] By adopting the above technical solution, it is possible to automatically calculate the priority coefficient according to the order urgency, customer level and contract constraint conditions, ensuring more reasonable resource allocation; in addition, by receiving order information in real time and dynamically adjusting the demand queue of the resource node according to the priority coefficient rule; this real-time adjustment can ensure that in the case of limited resources, it can quickly respond according to the actual situation, avoid resource conflicts, and maximize the utilization efficiency of resources.

[0102] A further setting is that the optimization operation types in S400 include resource node adjustment, association relationship correction and dynamic weight update;

[0103] The specific steps for generating the target optimization strategy set in S400 are as follows:

[0104] S410. If the optimization operation type is resource node adjustment, recalculate the dynamic weights of the affected nodes and update the resource reallocation plan;

[0105] S420. If the optimization operation type is the correction of the association relationship, adjust the association strength value, re-evaluate the critical link, and generate a new logistics path adjustment plan.

[0106] S430. If the optimization operation type is the dynamic update of weights, overwrite the original dynamic weights based on the direct weight value input by the user, and synchronously correct the priority adjustment plan.

[0107] By adopting the above technical solution, by introducing flexible optimization operation types, allowing users to select different optimization operations according to the actual situation, this flexibility enables the supply chain to continuously optimize in a dynamically changing environment, improving the adaptability and efficiency of the system.

[0108] A further setting is that the S500 specifically includes the following sub-steps:

[0109] S510. Execute the initial optimization strategy set after dynamic correction, including the resource reallocation plan, the logistics path adjustment plan, and the priority adjustment plan.

[0110] S520. Real-time collect the feedback data during the execution process. The feedback data includes the resource allocation completion rate, the actual transportation time of the logistics path, the order delivery timeliness, and the change value of the node resource utilization rate.

[0111] S530. Adjust the dynamic weights of the resource nodes in the supply chain map.

[0112] S540. Update the topological structure and the association strength value of the supply chain map based on the updated dynamic weights.

[0113] Among them, the specific implementation of adjusting the dynamic weights of the resource nodes in the supply chain map includes:

[0114] Based on the feedback data collected in S520, recalculate the historical supply-demand matching rate, the current demand priority coefficient, and the resource utilization rate of the resource nodes, and update the dynamic weights of each resource node:

[0115] If the resource allocation completion rate is lower than the set threshold, increase the value of the weight adjustment factor α to strengthen the influence of the historical supply-demand matching rate on the dynamic weight.

[0116] If the deviation of the logistics path transportation time exceeds the tolerance range, increase the value of the weight adjustment factor β to increase the weight proportion of the current demand priority coefficient.

[0117] If the change value of the resource utilization rate does not reach the target, increase the value of the weight adjustment factor γ to enhance the contribution degree of the resource utilization rate in the dynamic weight.

[0118] By adopting the above technical solutions, the weights of resource nodes can be dynamically adjusted according to the feedback data generated during the execution of the supply chain, further improving the supply chain map and making timely strategy adjustments. This closed-loop optimization mechanism based on feedback data ensures that the supply chain can continuously adapt to the changing environment and continuously enhance the optimization effect. First, by collecting real-time feedback data during the execution process, including the resource allocation completion rate, the actual transportation time of the logistics path, the order delivery timeliness, and the change in node resource utilization rate, these data reflect the effect of the optimization strategy during the actual execution process and can help the system identify which optimization measures have achieved the expected effect and which aspects need further adjustment. Then, based on the updated dynamic weights, the topological structure and the associated strength values of the supply chain map will also be updated accordingly. This process ensures that the supply chain map always reflects the real supply chain state, thus providing a more accurate basis for subsequent optimization decisions. The collection of feedback data and the update of the map enable the optimization process to have an adaptive ability, capable of making fine-tuning according to the actual execution situation and continuously improving the efficiency and response ability of the supply chain.

[0119] The present invention also provides a supply chain resource optimization system based on map analysis, including the following modules:

[0120] A supply chain map construction module for executing S100, including:

[0121] A data collection unit configured to collect historical supply chain data and real-time supply chain data;

[0122] A feature extraction unit configured to perform multi-dimensional feature extraction on the data to generate attribute information and dynamic weights of resource nodes;

[0123] A graph database unit configured to construct the topological structure of the supply chain map based on the graph database and bind node attributes and associated strength labels;

[0124] A map analysis engine module for executing S200, including:

[0125] A path traversal unit that uses the depth-first search algorithm to extract complete paths;

[0126] An efficiency calculation unit configured to calculate the resource flow efficiency of the path;

[0127] A bottleneck identification unit that screens key paths, inefficient paths, and confirms bottleneck nodes based on the efficiency threshold and dynamic weights;

[0128] An optimization strategy generation module for executing S300, including:

[0129] A resource reallocation unit that generates resource allocation instructions for inefficient paths;

[0130] Path adjustment unit, which generates alternative logistics paths for critical links;

[0131] Priority adjustment unit, which generates a priority adjustment plan for demand priority conflicts;

[0132] User interaction correction module, which is used to execute S400 and includes:

[0133] Signal reception interface, which receives the type of optimization operation input by the user;

[0134] Dynamic correction unit, which adjusts the initial optimization strategy set in real time according to the operation type;

[0135] Execution feedback and graph update module, which is used to execute S500 and includes:

[0136] Strategy execution unit, which is used to execute the resource reallocation plan, logistics path adjustment plan and priority adjustment plan;

[0137] Feedback collection unit, which monitors the execution results and extracts the resource allocation completion rate, logistics path transportation time deviation and utilization rate change value.

[0138] In summary, the present invention has the following beneficial effects:

[0139] It solves the key technical defects exposed by traditional supply chain optimization methods: unable to achieve dynamic global analysis and real-time response of the supply chain network; improving the efficiency of resource allocation and reducing the risk of logistics interruption in a complex supply chain environment. Description of the Drawings

[0140] Figure 1 It is a schematic flowchart of the embodiment;

[0141] Figure 2 It is a block diagram of the system structure of the embodiment. Detailed Embodiment

[0142] The present invention will be further described in detail below with reference to the accompanying drawings.

[0143] As shown in the attac Figure 1 hed drawings, this embodiment discloses a supply chain resource optimization method based on graph analysis, including the following steps:

[0144] S100. Construct a supply chain graph and update the supply chain graph based on real-time supply chain data; wherein, the supply chain graph includes multiple resource nodes, the association relationships between resource nodes, and the attribute information of each resource node, and the real-time supply chain data includes order information, inventory status, logistics dynamics, and external environment data;

[0145] S200. Identify critical paths, inefficient paths in the supply chain graph and confirm bottleneck nodes through graph analysis algorithms;

[0146] S300. Generate an initial optimization strategy set based on the critical path and bottleneck nodes; the initial optimization strategy set includes a resource reallocation plan, a logistics path adjustment plan, and a priority adjustment plan.

[0147] S400. Receive user interaction signals, determine the type of optimization operation according to the user interaction signals, and dynamically correct the initial optimization strategy set based on the type of optimization operation.

[0148] S500. Execute the initial optimization strategy set after dynamic correction, and update the supply chain map based on the execution results feedback.

[0149] Specifically, the specific implementation of constructing the supply chain map in S100 includes:

[0150] S110. Collect historical supply chain data, where the historical supply chain data includes supplier information, distribution channels, warehousing nodes, and customer demand records.

[0151] In this embodiment, the supply chain network of an electronic product manufacturer includes 5 suppliers, denoted as S1 - S5 respectively; 3 regional warehouses, denoted as W1 - W3 respectively; 2 distribution centers, denoted as D1 - D2 respectively; and an end - customer group, denoted as C1 - C100 respectively.

[0152] S120. Extract multi - dimensional features from the historical supply chain data to generate attribute information of resource nodes; the attribute information includes resource type, supply capacity, demand priority, resource utilization rate, historical supply - demand matching rate, and dynamic weight; where:

[0153] The demand priority is dynamically assigned according to the order urgency, customer level, and contract constraint conditions to obtain the corresponding demand priority coefficient.

[0154] In this embodiment, the resource type is marked as raw material, semi - finished product, or finished product node through material category and production equipment type.

[0155] The supply capacity is quantitatively calculated based on historical production capacity data, inventory turnover rate, and maximum load threshold.

[0156] The resource utilization rate is defined as the real - time ratio of the current resource usage to the total capacity.

[0157] The historical supply - demand matching rate is calculated based on historical order data to calculate the matching ratio of the actual delivery volume to the demand volume.

[0158] The dynamic weight is composed of the weighted sum of three parts, specifically including:

[0159] Divide the historical supply-demand matching rate of the resource node by the preset maximum matching rate benchmark value, and then multiply by the weight adjustment factor α to obtain the proportion of the historical supply-demand matching rate, which is used as the first part;

[0160] Divide the current demand priority coefficient by the average priority coefficient, and then multiply by the weight adjustment factor β to obtain the proportion of the current demand priority, which is used as the second part;

[0161] Divide the resource utilization rate by the total resource capacity, and then multiply by the weight adjustment factor γ to obtain the proportion of the resource utilization rate, which is used as the third part;

[0162] Add the first part, the second part, and the third part together to obtain the dynamic weight;

[0163] The dynamic weight W d The calculation formula is:

[0164] W d =α⋅R m / R max +β⋅P c / P avg +γ⋅U r / U total ;

[0165] Among them, W d is the dynamic weight, R m is the historical supply-demand matching rate, R max is the maximum matching rate benchmark value, P c is the current demand priority coefficient, P avg is the average priority coefficient, U r is the resource utilization rate, U total is the total resource capacity, and α, β, γ are weight adjustment factors and satisfy α + β + γ = 1.

[0166] In this embodiment, taking the warehouse node W2 as an example, calculate its dynamic weight:

[0167] The parameter settings are α = 0.4, β = 0.3, γ = 0.3, the maximum matching rate benchmark value = 95%, the average priority coefficient = 0.7, and the total resource capacity = 20,000 pieces;

[0168] The input data are respectively the historical supply-demand matching rate = 88%, the current demand priority coefficient = 0.9 (due to the urgent order of VIP customer C10), and the resource utilization rate = 75%;

[0169] The calculated value of the first part is 0.370, the value of the second part is 0.386, the value of the third part is 0.225. Add the first part, the second part, and the third part together to obtain the dynamic weight of 0.981.

[0170] S130. Construct the topological structure of the supply chain map based on the graph database; in the topological structure:

[0171] Resource nodes are connected by directed edges, and the directed edges represent the resource flow direction and the association strength;

[0172] Each resource node is bound with attribute information to form a complete node description including resource type tags, supply capacity values, demand priority coefficients, resource utilization rates, historical supply-demand matching rates, and dynamic weight parameters.

[0173] Specifically, the association strength in S130 is as follows:

[0174] S130.1. Based on the historical supply chain data collected in S110, count the resource interaction frequency and interaction volume between adjacent resource nodes;

[0175] In this embodiment, the resource interaction frequency represents the number of resource transfers between nodes per unit time calculated through the transaction records in the order information;

[0176] The resource interaction volume represents the total quantity or value of materials in a single interaction between adjacent nodes counted through the logistics dynamic data;

[0177] S130.2. Based on the resource interaction frequency, resource interaction volume, and node attribute information generated in S120, calculate the association strength value; the calculation process of the association strength value is as follows;

[0178] Divide the resource interaction frequency by the set maximum interaction frequency reference value to obtain the normalized interaction frequency ratio;

[0179] Divide the resource interaction volume by the set maximum interaction volume reference value to obtain the normalized interaction volume ratio;

[0180] Take the average value of the supply capacities of the source node and the target node as the contribution value of the supply capacity to the association strength;

[0181] Multiply the above three parts of the results by the preset weight coefficients respectively, and satisfy the condition that the sum of the weight coefficients is 1. The association strength value is the sum of the three parts;

[0182] S130.3. Assign weights to the directed edges according to the association strength value:

[0183] If the association strength value is greater than or equal to the first threshold, assign a high association strength label to this directed edge;

[0184] If the association strength value is between the second threshold and the first threshold, assign a medium association strength label to this directed edge;

[0185] If the association strength value is less than the second threshold, assign a low association strength label to this directed edge;

[0186] S130.4. Mark critical links based on the dynamic weight adjustment threshold in S140:

[0187] Re-adjust the first threshold according to the dynamic weights of resource nodes calculated in S140, so that the first threshold increases as the node importance increases;

[0188] Mark the association relationship as a critical link only when the association strength value reaches or exceeds the adjusted first threshold.

[0189] Specifically:

[0190] T1 = T BASE ⋅(1 + W d / W max ) ;

[0191] Wherein, T BASE is the first threshold before adjustment, W max is the set maximum dynamic weight value, W d is the dynamic weight, and T1 is the first threshold after adjustment.

[0192] Specifically, S200 specifically includes the following sub-steps:

[0193] S210. Traverse the supply chain graph using the depth-first search algorithm to extract all complete paths from suppliers to end customers; the traversal process includes the following sub-steps:

[0194] Select all initial nodes marked as suppliers from the supply chain graph and create independent search threads for each initial node;

[0195] Recursively visit adjacent nodes based on the depth-first search algorithm, and record the sequence of nodes passed through and the direction and strength of the associated edges;

[0196] If the current node is marked as an end customer node, confirm the recorded node sequence as a complete path, and save the dynamic weights of all nodes in the path and the weight labels of the associated edges;

[0197] Exclude paths containing loops or duplicate nodes to ensure that each path is an acyclic unidirectional link;

[0198] S220. Calculate the resource flow efficiency of each path;

[0199] S230. Mark critical paths and inefficient paths based on resource flow efficiency; specifically includes the following sub-steps:

[0200] Compare the resource flow efficiency value with the preset efficiency threshold interval, and the preset efficiency threshold interval includes a high-efficiency threshold and an inefficient threshold;

[0201] If the resource flow efficiency value is greater than or equal to the high - efficiency threshold, mark the corresponding path as the critical path;

[0202] If the resource flow efficiency value is less than the low - efficiency threshold, mark the corresponding path as the low - efficiency path;

[0203] S240. Perform differential analysis on the critical path and the low - efficiency path, and confirm the bottleneck nodes; specifically, it includes the following sub - steps:

[0204] Extract the associated edge with the highest association strength value in the critical path, mark it as the core link, and allocate additional resource redundancy to the core link, such as increasing inventory buffer or logistics alternative routes, to ensure its stability;

[0205] For each low - efficiency path, extract the dynamic weights of all nodes in the path, and screen out the node with the lowest dynamic weight; if the dynamic weights of multiple nodes are the same and are all the lowest values, further compare their historical supply - demand matching rates, and confirm the node with the lowest matching rate as the bottleneck node;

[0206] Associate and store the core link information of the critical path and the bottleneck node information of the low - efficiency path.

[0207] Specifically, the calculation of the resource flow efficiency in S220 includes the following sub - steps:

[0208] Extract the dynamic weights of all nodes in the path, and calculate the geometric mean of the node dynamic weights;

[0209] Extract the weight labels of all associated edges in the path, and calculate the arithmetic mean of the association strength values;

[0210] Statistical path physical length, where the physical length is defined as the sum of the geographical distances between adjacent nodes in the path;

[0211] Calculate the resource flow efficiency through the following comprehensive evaluation logic:

[0212] Multiply the geometric mean of the node dynamic weights by the arithmetic mean of the association strength values to obtain the comprehensive evaluation value;

[0213] Add the path physical length to a preset anti - zero constant, where the anti - zero constant is a very small positive number used to avoid calculation anomalies;

[0214] Divide the comprehensive evaluation value by the length adjustment value to obtain the resource flow efficiency value; the higher the resource flow efficiency value, the better the resource transfer ability and stability of the path.

[0215] In this embodiment, it is found that for the path S3→W2→D1→C10, the physical length = 350km.

[0216] The geometric mean of the dynamic weights of the nodes is calculated to be 0.84, the arithmetic mean of the association strength is calculated to be 0.82, the comprehensive evaluation value is calculated to be 0.69, and the resource flow efficiency is calculated to be 0.00197.

[0217] In the inefficient path S5→W3→D2→C25, the bottleneck node is W3.

[0218] Specifically, as shown in Table 1;

[0219] Table 1 Comparison of resource flow efficiency between critical paths and inefficient paths

[0220] Path Geometric mean of node dynamic weights Association strength (arithmetic mean) Physical length (km) Resource flow efficiency value Judgment result S3→W2→D1→C10 0.84 0.82 350 0.00197 Critical path S5→W3→D2→C25 0.62 0.58 420 0.00087 Inefficient path Preset threshold - - - High efficiency ≥ 0.0015 Inefficient < 0.001

[0221] Specifically, the specific implementation of generating the initial optimization strategy set in S300 includes:

[0222] S310. Generate a resource reallocation plan for the inefficient path; specifically including the following sub-steps:

[0223] Extract the inefficient paths marked in S230 and S240 and their corresponding bottleneck node information;

[0224] Traverse the adjacent nodes of the bottleneck node and filter out the adjacent nodes whose redundant resources meet the preset conditions; the filtering process is specifically that the current resource utilization rate of the adjacent node is lower than the preset utilization rate threshold, and its supply capacity is greater than the current load;

[0225] Calculate the resource gap of the bottleneck node, that is, the difference between the current demand and the current supply;

[0226] Based on the redundant resource capacity and association strength value of the adjacent node, allocate the resource demand according to the following rules:

[0227] Give priority to the adjacent nodes with high association strength values;

[0228] The allocation amount does not exceed the set percentage threshold of the redundant resources of the adjacent node, and the set percentage threshold is 80%;

[0229] Generate a resource transfer instruction and add the resource transfer instruction to the initial optimization strategy set;

[0230] In this embodiment, the demand gap of the bottleneck node W3 = 800 pieces, specifically the current inventory is 2,000 vs the demand is 2,800;

[0231] Adjacent node screening, the utilization rate of node W1 = 65%, specifically the redundant capacity = 7,000 pieces, and the association strength with node W3 = 0.76;

[0232] Transfer 500 pieces from node W1 and generate a transfer instruction.

[0233] S320. Generate a logistics path adjustment plan for the critical link, specifically including the following sub-steps:

[0234] Extract the critical path marked in S230 and its core link information;

[0235] Obtain the physical length of the current logistics path of the core link and external environment data;

[0236] If the physical length of the current logistics path exceeds the preset length threshold or the external environment risk level is higher than the preset risk threshold, perform the following adjustments:

[0237] Execute the path shortening strategy, specifically:

[0238] Through the topological structure of S130, search for links in the alternative path whose association strength value is not lower than that of the original path and whose physical length is shorter; if no alternative path can be found, retain the original path, but trigger a warning signal and generate a path optimization suggestion report, and the optimization suggestion report includes increasing logistics capacity, temporarily enabling standby nodes or adjusting the transportation time period;

[0239] Execute the risk avoidance strategy, specifically:

[0240] Based on the supplier information collected by S110 and the historical accident times and total transportation times recorded in the warehouse node data, calculate the risk coefficient of the nodes in the path; calculate the risk coefficient of the nodes in the path;

[0241] More specifically: the risk coefficient R = historical accident times / total transportation times × 100%;

[0242] Screen nodes with a risk coefficient lower than the preset threshold and adjacent in geographical location as standby nodes;

[0243] Replace the high-risk nodes with standby nodes, and recalculate the association strength value and physical length of the path after replacement;

[0244] Generate an optimized logistics path instruction and add the logistics path instruction to the initial optimization strategy set.

[0245] In this embodiment, due to the rainstorm warning for the original critical path S3→W2→D1:

[0246] Alternative path activated: S3→W1→D1 (association strength = 0.81, physical length = 320 km, risk coefficient reduced by 40%);

[0247] The system automatically switches the path and notifies the logistics fleet.

[0248] Specifically, the demand priority in S120 is dynamically assigned according to the order urgency, customer level and contract constraint conditions, and the process of obtaining the corresponding demand priority coefficient is specifically:

[0249] Divide the orders into four levels: urgent, high, medium, and low, and quantify the weight of the urgency level based on the contract constraint conditions;

[0250] Divide the customers into two categories: VIP and ordinary, and allocate the customer level weights according to the historical cooperation duration, order scale, and performance record;

[0251] Calculate the requirement priority coefficient through the following priority coefficient generation rules:

[0252] Allocate basic weights to the order urgency level and customer level respectively;

[0253] Weightedly sum the basic weights and the contract penalty ratio to generate the priority coefficient;

[0254] Link and adjust the priority coefficient and the node dynamic weight to ensure that high-priority resource nodes are preferentially matched with high-priority requirements;

[0255] The specific implementation of generating the initial optimization strategy set in S300 also includes:

[0256] S330. Generate a priority adjustment plan for requirement priority conflicts; specifically including the following sub-steps:

[0257] Receive the updated order information from S110 in real time, and parse the order urgency level and customer level;

[0258] Based on the priority coefficient generation rules defined in S120, recalculate the requirement priority coefficient of the conflict node;

[0259] If there are requirement conflicts for multiple orders on the same resource node, handle them according to the following rules:

[0260] Give priority to ensuring the requirements with a higher classification of order urgency level;

[0261] If the urgency level classifications are the same, determine the priority according to the customer level weight;

[0262] If the urgency level classifications and customer levels are both the same, sort them from high to low according to the contract penalty ratio;

[0263] Then, according to the updated priority coefficient, dynamically adjust the sorting of the requirement queues of the resource nodes:

[0264] Place the order with the highest priority coefficient at the top and allocate reserved resource capacity for it;

[0265] Generate a delay processing suggestion for the order with the second highest priority coefficient, including the estimated processing time and recommended alternative resource nodes;

[0266] Finally, add the adjusted priority instructions to the initial optimization strategy set.

[0267] In this embodiment, W2 receives the demands of C10 (VIP, urgent order) and C25 (ordinary, regular order) simultaneously; after calculation, the priority coefficient of C10 is 0.9, and the priority coefficient of C25 is 0.4;

[0268] The processing result is to reserve 1,000 pieces of resources for W2 to give priority to meeting C10, and the C25 order is postponed to the next day for processing.

[0269] Specifically, the optimization operation types in S400 include resource node adjustment, association relationship correction, and weight dynamic update;

[0270] Generating the target optimization strategy set in S400 specifically includes the following steps:

[0271] S410. If the optimization operation type is resource node adjustment, recalculate the dynamic weights of the affected nodes and update the resource reallocation plan;

[0272] S420. If the optimization operation type is association relationship correction, adjust the association strength value and re-evaluate the critical link to generate a new logistics path adjustment plan;

[0273] S430. If the optimization operation type is weight dynamic update, overwrite the original dynamic weight based on the direct weight value input by the user and synchronously correct the priority adjustment plan.

[0274] Specifically, S500 specifically includes the following sub-steps:

[0275] S510. Execute the initial optimization strategy set after dynamic correction, including the resource reallocation plan, the logistics path adjustment plan, and the priority adjustment plan;

[0276] S520. Real-time collect the feedback data during the execution process. The feedback data includes the resource allocation completion rate, the actual transportation time of the logistics path, the order delivery timeliness, and the change value of the node resource utilization rate;

[0277] In this embodiment, the process of collecting feedback data includes the following sub-steps:

[0278] Real-time monitor the execution status of the resource allocation instruction through the Internet of Things devices and the supply chain management system, record the difference between the actual resource allocation quantity and the planned allocation quantity, and calculate the resource allocation completion rate;

[0279] Obtain the GPS trajectory data after the logistics path is adjusted, count the deviation between the actual transportation time and the expected transportation time, and record the occurrence times of external environmental risk events;

[0280] Analyze the log data in the order processing system, extract the total duration from order submission to delivery, and calculate the improvement rate of order delivery timeliness;

[0281] Collect the real-time resource usage of each resource node, recalculate the resource utilization rate in combination with the total resource capacity, compare it with the resource utilization rate before optimization, and generate the change value of the resource utilization rate.

[0282] S530. Adjust the dynamic weights of resource nodes in the supply chain map;

[0283] S540. Update the topological structure and association strength value of the supply chain map based on the updated dynamic weights;

[0284] Among them, the specific implementation of adjusting the dynamic weights of resource nodes in the supply chain map includes:

[0285] Based on the feedback data collected in S520, recalculate the historical supply-demand matching rate, current demand priority coefficient, and resource utilization rate of resource nodes, and update the dynamic weights of each resource node:

[0286] If the resource allocation completion rate is lower than the set threshold, increase the value of the weight adjustment factor α to strengthen the influence of the historical supply-demand matching rate on the dynamic weight;

[0287] If the deviation of the logistics path transportation time exceeds the tolerance range, increase the value of the weight adjustment factor β to increase the weight ratio of the current demand priority coefficient;

[0288] If the change value of the resource utilization rate does not reach the target, increase the value of the weight adjustment factor γ to enhance the contribution degree of the resource utilization rate in the dynamic weight.

[0289] In this embodiment, the actual completion time of the W1→W3 allocation is 2 hours later than expected (the original plan was 4 hours), resulting in the utilization rate of W3 rising to 82%;

[0290] Due to the transportation time deviation, the system automatically increases β from 0.3 to 0.35 to enhance the influence of the demand priority in the dynamic weight;

[0291] The dynamic weight of W3 is recalculated to change from the original 0.58 to 0.92, triggering the intensity label of the associated edge S5→W3 to be upgraded from the low association intensity label to the medium association intensity label.

[0292] As shown in the appendix Figure 2 This embodiment also discloses a supply chain resource optimization system based on graph analysis, including the following modules:

[0293] The supply chain map construction module is used to execute S100, including:

[0294] A data acquisition unit, configured to acquire historical supply chain data and real-time supply chain data;

[0295] A feature extraction unit, configured to perform multi-dimensional feature extraction on the data to generate attribute information and dynamic weights of resource nodes;

[0296] A graph database unit, configured to construct the topological structure of the supply chain map based on the graph database, and bind node attributes and association strength labels;

[0297] A map analysis engine module, used to execute S200, including:

[0298] A path traversal unit, which uses the depth-first search algorithm to extract the complete path;

[0299] An efficiency calculation unit, configured to calculate the resource flow efficiency of the path;

[0300] A bottleneck identification unit, which screens critical paths, inefficient paths and confirms bottleneck nodes based on the efficiency threshold and dynamic weights;

[0301] An optimization strategy generation module, used to execute S300, including:

[0302] A resource reallocation unit, which generates a resource transfer instruction for the inefficient path;

[0303] A path adjustment unit, which generates an alternative logistics path plan for the critical link;

[0304] A priority adjustment unit, which generates a priority adjustment plan for demand priority conflicts;

[0305] A user interaction correction module, used to execute S400, including:

[0306] A signal receiving interface, which receives the type of optimization operation input by the user;

[0307] A dynamic correction unit, which adjusts the initial optimization strategy set in real time according to the operation type;

[0308] An execution feedback and map update module, used to execute S500, including:

[0309] A strategy execution unit, used to execute the resource reallocation plan, the logistics path adjustment plan and the priority adjustment plan;

[0310] A feedback collection unit, which monitors the execution result and extracts the resource transfer completion rate, the deviation of the logistics path transportation time and the change value of the utilization rate.

[0311] This specific embodiment is only an interpretation of the present invention and does not limit the present invention. After reading this specification, those skilled in the art can make modifications to this embodiment without creative contributions as needed, but as long as it is within the scope of the claims of the present invention, it is protected by the patent law.

Claims

1. A supply chain resource optimization method based on graph analysis, characterized in that: The following steps are involved: S100, constructing a supply chain graph, and updating the supply chain graph based on real-time supply chain data; wherein the supply chain graph includes multiple resource nodes, associations between resource nodes, and attribute information of each resource node, and the real-time supply chain data includes order information, inventory status, logistics dynamics, and external environment data; S200, identify the critical paths and inefficient paths in the supply chain graph and confirm the bottleneck nodes through graph analysis algorithms; S300, generating an initial optimization strategy set based on the critical path and bottleneck nodes; the initial optimization strategy set includes a resource reallocation plan, a logistics path adjustment plan, and a priority adjustment plan; S400, receiving a user interaction signal, and determining an optimization operation type according to the user interaction signal, and dynamically modifying the initial optimization strategy set based on the optimization operation type; S500, executing the dynamically revised initial optimization strategy set, and updating the supply chain map based on the feedback of the execution result; Wherein, the S300 includes: S310, generating a resource reallocation plan for an inefficient path: Extract the marked inefficient paths and their corresponding bottleneck node information; Traverse the adjacent nodes of the bottleneck node and select the adjacent nodes whose redundant resources meet the preset conditions; the preset conditions are that the current resource utilization of the adjacent nodes is lower than the preset utilization threshold and its supply capacity is greater than the current load; Calculate the resource gap of the bottleneck node, that is, the difference between the current demand and the current supply; Based on the redundant resource capacity and association strength values ​​of adjacent nodes, resource requirements are allocated according to the following rules: Prioritize adjacent nodes with high association strength values; The allocated amount does not exceed the set percentage threshold of the redundant resources of the adjacent nodes; Generate a resource allocation instruction, and add the resource allocation instruction to the initial optimization strategy set; S320. Generate logistics route adjustment plan for key links: Extract the marked critical path and its core link information; Obtain the current physical length of the logistics path and external environment data of the core link; If the physical length of the current logistics path exceeds the preset length threshold or the external environment risk level is higher than the preset risk threshold, the following adjustments are performed: Execute the path shortening strategy, specifically: Through the topological structure of the supply chain graph, search for links in the alternative path whose association strength value is not lower than that of the original path and whose physical length is shorter; if no alternative path can be found, retain the original path, but trigger an early warning signal and generate a path optimization recommendation report, which includes increasing logistics capacity, temporarily enabling spare nodes, or adjusting transportation time periods; Execute risk avoidance strategies, specifically: Calculate the risk coefficient of the nodes in the path based on the collected supplier information and the number of historical accidents and total transportation times recorded in the warehouse node data; Select nodes with risk factors lower than the preset threshold and geographically adjacent as backup nodes; Replace high-risk nodes with spare nodes, and recalculate the association strength value and physical length of the replaced path; An optimized logistics path instruction is generated, and the logistics path instruction is added to the initial optimization strategy set.

2. The supply chain resource optimization method based on graph analysis according to claim 1 is characterized in that: The specific implementation of constructing the supply chain map in S100 includes: S110, collecting historical supply chain data, where the historical supply chain data includes supplier information, distribution channels, storage nodes, and customer demand records; S120, extract multi-dimensional features from the historical supply chain data to generate attribute information of resource nodes; the attribute information includes resource type, supply capacity, demand priority, resource utilization, historical supply and demand matching rate and dynamic weight; wherein: The demand priority is dynamically assigned according to the order urgency, customer level and contract constraints to obtain the corresponding demand priority coefficient; The dynamic weight is composed of the weighted sum of three parts, specifically including: Divide the historical supply-demand matching rate of the resource node by the preset maximum matching rate benchmark value, and then multiply it by the weight adjustment factor α to obtain the historical supply-demand matching rate ratio as the first part; Divide the current demand priority coefficient by the average priority coefficient, and then multiply by the weight adjustment factor β to get the current demand priority ratio as the second part; Divide the resource utilization by the total resource capacity, and then multiply by the weight adjustment factor γ to get the resource utilization ratio as the third part; Adding the first part, the second part and the third part to obtain the dynamic weight; S130. Construct a topological structure of the supply chain graph based on a graph database; in the topological structure: Resource nodes are connected by directed edges, which represent the direction of resource flow and the strength of association; Each resource node is bound to the attribute information to form a complete node description including resource type label, supply capacity value, demand priority coefficient, resource utilization rate, historical supply and demand matching rate and dynamic weight parameters.

3. The supply chain resource optimization method based on graph analysis according to claim 2 is characterized in that: The association strength in S130 is specifically: S130.

1. Based on the historical supply chain data collected by S110, count the resource interaction frequency and interaction volume between adjacent resource nodes; S130.

2. Calculate the association strength value based on the resource interaction frequency, the resource interaction amount and the node attribute information generated in S120. The calculation process of the association strength value is as follows; Divide the resource interaction frequency by the set maximum interaction frequency benchmark value to obtain the normalized interaction frequency ratio; Divide the resource interaction volume by the set maximum interaction volume benchmark value to obtain the normalized interaction volume ratio; The average value of the supply capacity of the source node and the target node is taken as the contribution value of the supply capacity to the association strength; The normalized interaction frequency ratio, the normalized interaction volume ratio and the contribution value of the supply capacity to the association strength are multiplied by the preset weight coefficients to obtain the association strength value, satisfying the condition that the sum of the weight coefficients is 1; S130.

3. Assign weights to directed edges based on the association strength value: If the association strength value is greater than or equal to the first threshold, assigning a high association strength label to the directed edge; If the association strength value is between the second threshold and the first threshold, a medium association strength label is assigned to the directed edge; If the association strength value is less than the second threshold, a low association strength label is assigned to the directed edge; S130.

4. Based on the dynamic weight adjustment threshold in S140, mark the key links: According to the dynamic weight of the resource node calculated in S140, the first threshold is readjusted so that the first threshold increases as the importance of the node increases; Only when the association strength value reaches or exceeds the adjusted first threshold value, the association relationship is marked as a critical link.

4. The supply chain resource optimization method based on graph analysis according to claim 2 is characterized in that: The S200 specifically includes the following sub-steps: S210, using a depth-first search algorithm to traverse the supply chain graph and extract all complete paths from suppliers to end customers; S220, calculating the resource flow efficiency of each path; S230, marking critical paths and inefficient paths based on resource flow efficiency; It includes the following sub-steps: Comparing the resource flow efficiency value with a preset efficiency threshold interval, where the preset efficiency threshold interval includes a high efficiency threshold and a low efficiency threshold; If the resource flow efficiency value is greater than or equal to the high efficiency threshold, the corresponding path is marked as a critical path; If the resource flow efficiency value is less than the inefficiency threshold, the corresponding path is marked as an inefficient path; S240, perform differential analysis on critical paths and inefficient paths, and identify bottleneck nodes; It includes the following sub-steps: Extract the associated edge with the highest association strength value in the critical path, mark it as the core link, and allocate additional resource redundancy to the core link to ensure its stability; For each inefficient path, extract the dynamic weights of all nodes in the path and select the node with the lowest dynamic weight. If the dynamic weights of multiple nodes are the same and are the lowest, further compare their historical supply-demand matching rates and identify the node with the lowest matching rate as the bottleneck node. The core link information of the critical path and the bottleneck node information of the inefficient path are associated and stored.

5. The supply chain resource optimization method based on graph analysis according to claim 4 is characterized in that: The calculation of resource flow efficiency in S220 includes the following sub-steps: Extract the dynamic weights of all nodes in the path and calculate the geometric mean of the dynamic weights of the nodes; Extract the weight labels of all associated edges in the path and calculate the arithmetic mean of the associated strength values; Counting the physical length of the path, where the physical length is defined as the sum of the geographical distances between adjacent nodes in the path; The resource flow efficiency is calculated by the following comprehensive evaluation logic: Multiply the geometric mean of the node dynamic weights by the arithmetic mean of the association strength values ​​to obtain a comprehensive evaluation value; Adding the physical length of the path to a preset anti-zero constant, where the anti-zero constant is a very small positive number for avoiding calculation anomalies; The comprehensive evaluation value is divided by the length adjustment value to obtain a resource flow efficiency value; the higher the resource flow efficiency value is, the better the resource transfer capacity and stability of the path are.

6. The supply chain resource optimization method based on graph analysis according to claim 2 is characterized in that: In S120, the demand priority is dynamically assigned according to the order urgency, customer level and contract constraints, and the corresponding demand priority coefficient is obtained as follows: Classify orders into four levels: urgent, high, medium, and low, and quantify the urgency weight based on contract constraints; Divide customers into two categories: VIP and ordinary, and assign customer level weights based on historical cooperation duration, order size, and fulfillment record; The demand priority coefficient is calculated by the following priority coefficient generation rule: Assign basic weights to order urgency and customer level respectively; The basic weight and the contract penalty ratio are weighted and summed to generate the priority coefficient; The priority coefficient is adjusted in conjunction with the node dynamic weight to ensure that high-priority resource nodes are matched with high-priority requirements first; The specific implementation of generating the initial optimization strategy set in S300 also includes: S330, generating a priority adjustment plan for demand priority conflicts; specifically comprising the following sub-steps: Receive the order information updated by S110 in real time, and analyze the order urgency and customer level; Recalculate the demand priority coefficients of the conflicting nodes based on the priority coefficient generation rule defined in S120; If there are demand conflicts between multiple orders on the same resource node, the following rules will be followed: Prioritize the orders with higher urgency level; If the urgency classification is the same, the priority is determined based on the customer level weight; If the urgency classification and customer level are the same, they will be sorted from high to low according to the contract penalty ratio; Next, according to the updated priority coefficient, dynamically adjust the demand queue order of the resource node: The orders with the highest priority coefficient are placed at the top and reserved resource capacity is allocated to them; Generate delayed processing suggestions for orders with the second highest priority coefficient, including estimated processing time and alternative resource node recommendations; Finally, the adjusted priority instructions are added to the initial optimization strategy set.

7. The supply chain resource optimization method based on graph analysis according to claim 2 is characterized in that: The optimization operation types in S400 include resource node adjustment, association relationship correction and weight dynamic update; The generating of the target optimization strategy set in S400 specifically includes the following steps: S410: If the optimization operation type is resource node adjustment, recalculate the dynamic weights of the affected nodes and update the resource reallocation plan; S420, if the optimization operation type is association relationship modification, then adjust the association strength value and re-evaluate the key links to generate a new logistics path adjustment plan; S430: If the optimization operation type is dynamic weight update, the original dynamic weight is overwritten based on the direct weight value input by the user, and the priority adjustment scheme is modified synchronously.

8. The supply chain resource optimization method based on graph analysis according to claim 2 is characterized by: The S500 specifically includes the following sub-steps: S510, executing the dynamically revised initial optimization strategy set, including a resource reallocation plan, a logistics path adjustment plan, and a priority adjustment plan; S520, collecting feedback data during the execution process in real time, wherein the feedback data includes resource allocation completion rate, actual transportation time of logistics path, order delivery timeliness and node resource utilization rate change value; S530, adjusting the dynamic weights of resource nodes in the supply chain graph; S540, updating the topological structure and association strength value of the supply chain graph based on the updated dynamic weight; The specific implementation of adjusting the dynamic weight of resource nodes in the supply chain graph includes: Based on the feedback data collected by S520, the historical supply-demand matching rate, current demand priority coefficient and resource utilization rate of resource nodes are recalculated, and the dynamic weight of each resource node is updated: If the resource allocation completion rate is lower than the set threshold, the value of the weight adjustment factor α is increased to strengthen the impact of the historical supply and demand matching rate on the dynamic weight; If the logistics path transportation time deviation exceeds the tolerance range, the value of the weight adjustment factor β is increased to increase the weight ratio of the current demand priority coefficient; If the resource utilization change value does not reach the target, the value of the weight adjustment factor γ is increased to enhance the contribution of resource utilization in the dynamic weight.

9. A supply chain resource optimization system based on graph analysis, applied to a supply chain resource optimization method based on graph analysis as claimed in any one of claims 1 to 8, characterized in that: Includes the following modules: Supply chain graph building blocks for executing S100 include: a data collection unit configured to collect historical supply chain data and real-time supply chain data; A feature extraction unit configured to extract multi-dimensional features from data and generate attribute information and dynamic weights of resource nodes; A graph database unit, configured to construct a topological structure of a supply chain graph based on a graph database, and to bind node attributes and association strength labels; The graph analysis engine module is used to execute S200, including: The path traversal unit uses a depth-first search algorithm to extract the complete path; an efficiency calculation unit configured to calculate resource flow efficiency of a path; Bottleneck identification unit, which screens critical paths, inefficient paths and identifies bottleneck nodes based on efficiency thresholds and dynamic weights; The optimization strategy generation module is used to execute S300, including: A resource reallocation unit generates resource allocation instructions for inefficient paths; The route adjustment unit generates logistics route alternatives for key links; A priority adjustment unit, generating a priority adjustment plan for demand priority conflicts; The user interaction correction module is used to execute S400, including: A signal receiving interface for receiving the optimization operation type input by the user; Dynamic correction unit, which adjusts the initial optimization strategy set in real time according to the operation type; The execution feedback and graph update module is used to execute S500, including: Strategy execution unit, used to execute resource reallocation plan, logistics path adjustment plan and priority adjustment plan; The feedback collection unit monitors the execution results and extracts the resource allocation completion rate, logistics path transportation time deviation and utilization change value.

Images