An industry chain core advantage research and strengthening method based on multi-agent cooperation

Abstract

The application relates to the technical field of artificial intelligence, and discloses an industry chain core advantage judgment and reinforcement method based on multi-agent cooperation, which comprises the following steps: collecting node operation data of each agent node in an industry chain network, calculating a cooperative operation index corresponding to the agent node; obtaining historical interaction records of the industry chain network, extracting associated features in the historical interaction records to analyze the advantage transmission efficiency between the agent nodes; calculating the output gap of the industry chain network, setting the node contribution rank of the agent node in the industry chain, calculating the total support amount of the core resource system, calculating the maximum carrying limit of the industry chain network at the current stage, and determining the available resource ratio corresponding to the industry chain network; and performing configuration reinforcement processing on the industry chain network to obtain a reinforcement result. The application can improve the efficiency of the multi-agent cooperative industry chain core advantage judgment and reinforcement.

Description

Technical Field

[0001] This invention relates to a method for assessing and strengthening the core advantages of an industrial chain based on multi-agent collaboration, belonging to the field of artificial intelligence technology. Background Technology

[0002] Supply chain synergy optimization is a crucial means to enhance regional economic and industrial competitiveness, and its efficiency management has a significant impact on achieving high-quality development goals. With the widespread application of digital and intelligent technologies in industrial governance, it is therefore necessary to accurately assess and dynamically strengthen the core advantages of multi-agent collaboration within the supply chain.

[0003] Existing supply chain assessment methods mostly employ basic analysis based on static data and isolated indicators, evaluating individual node enterprises through financial reports or periodic surveys. However, these methods fail to comprehensively consider the dynamic interaction effects between nodes, historical collaboration status, and the continuous differences between actual output levels and target requirements. Furthermore, they cannot coordinate with the real-time supply capacity and reserve status of the core resource system, which can easily lead to misallocation and efficiency losses during periods of abundant resources, thus limiting the improvement of the overall competitive advantage of the supply chain. Therefore, a method is needed that can improve the efficiency of judging and strengthening the core advantages of the supply chain through multi-agent collaboration. Summary of the Invention

[0004] This invention provides a method for judging and strengthening the core advantages of the industrial chain based on multi-agent collaboration, the main purpose of which is to improve the efficiency of judging and strengthening the core advantages of the industrial chain based on multi-agent collaboration.

[0005] To achieve the above objectives, this invention provides a method for assessing and strengthening core advantages in the industrial chain based on multi-agent collaboration, comprising:

[0006] Collect node operation data of each intelligent agent node in the industrial chain network, and calculate the collaborative operation index corresponding to the intelligent agent node based on the node operation data;

[0007] Obtain historical interaction records of the industry chain network, extract the correlation features in the historical interaction records, and analyze the efficiency of advantage transfer between the intelligent agent nodes;

[0008] The current output level and target output requirement of the industrial chain network are queried to calculate the output gap of the industrial chain network. Based on the collaborative operation index, the advantage transmission efficiency and the output gap, the node contribution position of the intelligent agent node in the industrial chain is set.

[0009] The real-time supply capacity and reserve level of the core resource system on which the industrial chain network depends are obtained in order to calculate the total support of the core resource system. Based on the total support, the maximum carrying capacity of the industrial chain network at the current stage is calculated to determine the available resource allocation corresponding to the industrial chain network.

[0010] By combining the node contribution ranking and the available resource allocation, a configuration enhancement process is performed on the industry chain network to obtain the enhancement result.

[0011] Optionally, calculating the cooperative operation index corresponding to the agent node based on the node's operational data includes:

[0012] The node's operational data is subjected to quality screening to obtain valid operational data;

[0013] Node interaction features are extracted from the valid operational data, and the node interaction features are normalized to obtain standard interaction features.

[0014] Correlation analysis was performed on the standard interaction features to obtain key collaborative features;

[0015] Based on the aforementioned key collaborative features, a node collaborative relationship graph is constructed;

[0016] Based on the node collaboration relationship diagram and the key collaboration features, the collaborative operation index corresponding to the agent node is calculated.

[0017] Optionally, the step of performing correlation analysis on the standard interaction features to obtain key collaborative features includes:

[0018] The standard interaction features are processed into feature labels to obtain interaction feature labels;

[0019] Calculate the association strength value between the interactive feature labels;

[0020] Based on the association strength value, the collaborative interaction features in the standard interaction features are extracted;

[0021] Calculate the feature redundancy corresponding to the collaborative interaction features;

[0022] Based on the aforementioned feature redundancy, the collaborative interaction features are filtered to obtain key collaborative features.

[0023] Optionally, calculating the cooperative operation index corresponding to the agent node by combining the node cooperative relationship graph and the key cooperative features includes:

[0024] The node collaboration graph is subjected to graph embedding processing to obtain the node structure feature vector;

[0025] The key collaborative features are reconstructed to obtain node behavior feature vectors;

[0026] Calculate the structural similarity between the node structural feature vectors, and determine the structural coordination index corresponding to the agent node based on the structural similarity;

[0027] Calculate the behavior matching degree between the node behavior feature vectors, and determine the behavior coordination index corresponding to the agent node based on the behavior matching degree;

[0028] By combining the structural coordination index and the behavioral coordination index, the coordinated operation index corresponding to the agent node is determined.

[0029] Optionally, extracting the association features from the historical interaction records includes:

[0030] Identify the interaction record text in the historical interaction records and calculate the text information entropy corresponding to the interaction record text;

[0031] Based on the text information entropy, extract the core interaction records from the historical interaction records;

[0032] The core interaction records are processed in a structured manner to obtain field-structured interaction records;

[0033] Extract the interaction association factors from the field structure interaction records, perform association pattern recognition on the interaction association factors, and obtain association pattern features;

[0034] The associated pattern features are subjected to feature aggregation processing to obtain associated features.

[0035] Optionally, analyzing the advantage transfer efficiency between the agent nodes based on the associated features includes:

[0036] The efficiency-related features corresponding to the agent nodes are filtered out from the associated features, and the real-time interaction data of the agent nodes are obtained.

[0037] Calculate the transmission efficiency index corresponding to the efficiency-related characteristics;

[0038] Based on the real-time interaction data, it is determined whether there is an interaction interruption between the intelligent agent nodes. If there is no interaction interruption, the superior transmission efficiency between the intelligent agent nodes is analyzed by combining the transmission efficiency index and the efficiency-related characteristics.

[0039] If an interaction interruption occurs, the interruption information related to the interaction interruption is extracted from the real-time interaction data, and the interruption impact level corresponding to the interaction interruption is analyzed based on the interruption information.

[0040] By combining the interruption impact level and the transmission efficiency index, the superior transmission efficiency between the agent nodes is analyzed.

[0041] Optionally, calculating the transmission efficiency index corresponding to the efficiency-related feature includes:

[0042] Extract multi-source collaborative features corresponding to efficiency-related features, and determine the collaborative efficiency value corresponding to the efficiency-related features based on the multi-source collaborative features;

[0043] Extract the dynamic evolution characteristics of the efficiency-related features, analyze the transmission contribution of the dynamic evolution characteristics in advantage assessment, and calculate the evolutionary stability corresponding to the efficiency-related features.

[0044] Based on the dynamic evolution characteristics, assign evolution weight coefficients corresponding to the efficiency-related characteristics;

[0045] The current assessment requirements for the core advantages of the industrial chain are obtained, and based on the assessment requirements, the transfer adjustment margin and noise tolerance threshold of the efficiency-related characteristics are calculated.

[0046] By combining the cooperative efficiency value, the evolution weight coefficient, the transmission adjustment margin, the evolution stability, and the noise tolerance threshold, the transmission efficiency index corresponding to the efficiency-related feature is calculated.

[0047] Optionally, the step of querying the current output level and target output requirement of the industry chain network to calculate the output gap of the industry chain network includes:

[0048] Extract the effective output data from the current output level;

[0049] Identify key gap elements and their corresponding performance values ​​from the effective output data, and obtain the element calibration values ​​corresponding to the key gap elements.

[0050] Calculate the element sensitivity coefficients corresponding to the key gap elements;

[0051] Based on the sensitivity coefficient of the factor, the performance value of the factor, and the calibration value of the factor, the output gap of the industrial chain network is calculated.

[0052] Optionally, calculating the element sensitivity coefficient corresponding to the key gap element includes:

[0053] Independent component separation is performed on the key gap elements to obtain the element eigenmodes;

[0054] Analyze the correlation mapping relationship between the intrinsic modes of the elements;

[0055] Based on the aforementioned association mapping relationship, construct the element action path corresponding to the key gap element;

[0056] Based on the action path of the aforementioned elements, the node action intensity corresponding to the key gap elements is calculated.

[0057] Based on the node effect strength, calculate the element sensitivity coefficient corresponding to the key gap element.

[0058] Optionally, the step of obtaining the real-time supply capacity and reserve level of the core resource system upon which the industrial chain network depends, in order to calculate the overall support capacity of the core resource system, includes:

[0059] The real-time supply capacity is stabilized to obtain the continuous supply level;

[0060] The reserve level is divided into time periods to obtain the stock distribution sequence;

[0061] The system tolerance is obtained by performing a collaborative matching analysis between the continuous supply level and the stock distribution sequence.

[0062] The system's tolerance is comprehensively calibrated to obtain the effective support capacity;

[0063] Based on the actual effective support capacity and the preset operating requirement cycle, the total support capacity of the core resource system is calculated.

[0064] To address the aforementioned problems, this invention also provides a system for assessing and strengthening core advantages in the industrial chain based on multi-agent collaboration, the system comprising:

[0065] The collaborative operation index calculation module is used to collect node operation data of each intelligent agent node in the industrial chain network, and calculate the collaborative operation index corresponding to the intelligent agent node based on the node operation data.

[0066] The advantage transfer efficiency analysis module is used to obtain historical interaction records of the industrial chain network, extract the correlation features in the historical interaction records, and analyze the advantage transfer efficiency between the intelligent agent nodes.

[0067] The node contribution ranking module is used to query the current output level and target output requirements of the industrial chain network to calculate the output gap of the industrial chain network. Combining the collaborative operation index, the advantage transmission efficiency and the output gap, the module sets the node contribution ranking of the intelligent agent node in the industrial chain.

[0068] The available resource allocation determination module is used to obtain the real-time supply capacity and reserve level of the core resource system on which the industrial chain network depends, so as to calculate the total support of the core resource system, and based on the total support, calculate the maximum carrying capacity of the industrial chain network at the current stage, so as to determine the available resource allocation corresponding to the industrial chain network.

[0069] A configuration enhancement processing module is configured to combine the node contribution ranking and the available resource allocation to perform configuration enhancement processing on the industry chain network and obtain enhancement results.

[0070] Compared to the problems described in the background art, this invention, by collecting node operation data of each intelligent agent node in the industrial chain network and calculating the corresponding collaborative operation index based on the node operation data, can grasp the collaborative performance of different nodes in the industrial chain, providing a data foundation for subsequent evaluation of the overall operational efficiency of the industrial chain. Furthermore, by acquiring historical interaction records of the industrial chain network and extracting correlation features, this invention can understand the connection strength and resource flow patterns of intelligent agent nodes in historical interactions, providing a basis for analyzing the efficiency of advantage transfer between nodes. Moreover, by querying the current output level and target output requirements of the industrial chain network, this invention can calculate the output gap of the industrial chain network, accurately quantifying the actual operation of the industrial chain network relative to the expected output target. The degree of deviation objectively reflects the achievement of production efficiency standards, providing a key basis for subsequently setting the node contribution ranking of the intelligent agent nodes in the industrial chain. Furthermore, by acquiring the real-time supply capacity and reserve level of the core resource system upon which the industrial chain network depends, this invention calculates the overall support capacity of the core resource system, accurately grasping the supply status of various key resources, avoiding industrial chain operation interruptions due to insufficient resource supply, and providing a reliable basis for subsequently calculating the maximum carrying capacity of the industrial chain network at the current stage. Finally, by combining the node contribution ranking and the available resource allocation ratio, this invention performs configuration enhancement processing on the industrial chain network, improving the accuracy and systematicness of industrial chain resource allocation, and achieving optimization and improvement of the overall efficiency of the industrial chain. Therefore, the method for judging and strengthening the core advantages of the industrial chain based on multi-agent collaboration provided by this invention can improve the efficiency of judging and strengthening the core advantages of the industrial chain based on multi-agent collaboration. Attached Figure Description

[0071] Figure 1 A flowchart illustrating a method for assessing and strengthening core advantages in the industrial chain based on multi-agent collaboration, provided in an embodiment of the present invention;

[0072] Figure 2 A schematic diagram of the configuration enhancement process for a method for judging and enhancing the core advantages of the industrial chain based on multi-agent collaboration, provided in an embodiment of the present invention;

[0073] Figure 3 This is a schematic diagram of a module of a system for judging and strengthening the core advantages of the industrial chain based on multi-agent collaboration, provided in an embodiment of the present invention.

[0074] The objectives, features, and advantages of this invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0075] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0076] This application provides a method for assessing and strengthening core advantages in the industrial chain based on multi-agent collaboration. The executing entity of this method includes, but is not limited to, at least one of the following electronic devices that can be configured to execute the method provided in this application: a server, a terminal, etc. In other words, this method can be executed by software or hardware installed on a terminal device or a server device. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster.

[0077] Reference Figure 1 The diagram shown is a flowchart illustrating a method for assessing and strengthening core advantages in the industrial chain based on multi-agent collaboration, according to an embodiment of the present invention. In this embodiment, the method for assessing and strengthening core advantages in the industrial chain based on multi-agent collaboration includes:

[0078] S1. Collect node operation data of each intelligent agent node in the industrial chain network, and calculate the collaborative operation index corresponding to the intelligent agent node based on the node operation data.

[0079] This invention collects node operation data from each intelligent agent node in the industrial chain network, and calculates the collaborative operation index corresponding to the intelligent agent node based on the node operation data. This allows us to understand the collaborative performance of different nodes in the industrial chain and provides a data foundation for subsequent evaluation of the overall operational efficiency of the industrial chain.

[0080] The industrial chain network refers to an industrial collaboration system composed of multiple interconnected intelligent agent nodes. Each intelligent agent node is an independent unit within the industrial chain with autonomous operation capabilities. Node operation data refers to various real-time records generated by the intelligent agent nodes during operation, including real-time information on production progress, resource consumption, interaction frequency, response time, and other aspects. This information reflecting the node operation process constitutes the node operation data. The collaborative operation index is a comprehensive quantitative evaluation of the collaborative efficiency, resource coordination capabilities, and interaction effects of each intelligent agent node in the industrial chain network. Furthermore, node operation data of each intelligent agent node in the industrial chain network can be collected through a combination of intelligent sensing terminals and system interfaces. Production sensors and equipment condition sensors are deployed for production-type intelligent agent nodes; for logistics-type intelligent agent nodes, GPS positioning terminals and loading / unloading barcode scanning equipment are deployed. Simultaneously, API interfaces are used to connect to the node's production management system and warehouse management system to achieve real-time data capture and historical data retrieval.

[0081] As an embodiment of the present invention, the step of calculating the cooperative operation index corresponding to the agent node based on the node operation data includes:

[0082] The node's operational data is subjected to quality screening to obtain valid operational data;

[0083] Node interaction features are extracted from the valid operational data, and the node interaction features are normalized to obtain standard interaction features.

[0084] Correlation analysis was performed on the standard interaction features to obtain key collaborative features;

[0085] Based on the aforementioned key collaborative features, a node collaborative relationship graph is constructed;

[0086] Based on the node collaboration relationship diagram and the key collaboration features, the collaborative operation index corresponding to the agent node is calculated.

[0087] The effective operational data refers to reliable data obtained after the node operational data has undergone quality screening operations such as removing abnormal records and correcting error information; the node interaction features are multiple dimensions of information reflecting the collaborative relationship between nodes in the effective operational data, such as interaction frequency, response time, and resource sharing degree; the standard interaction features are uniform dimensional data formed by standardizing the node interaction features; the key collaborative features are a subset of core features in the standard interaction features that are highly correlated with the collaborative operation index; and the node collaborative relationship graph is a network structure describing the collaborative relationship between nodes constructed based on the key collaborative features.

[0088] Furthermore, the node operation data can be quality-screened using the interquartile range (ICM) detection method to identify and process abnormal fluctuation data, thereby obtaining valid operation data. Node interaction features can be extracted from the valid operation data using feature engineering methods, including calculating the number of interactions between nodes per unit time and the average response time. The node interaction features can be normalized using the Z-score normalization method to eliminate differences in the magnitude of features, resulting in standard interaction features. Finally, a node collaboration graph can be established based on the key collaboration features using graph construction techniques, where nodes represent agent nodes and edge weights represent collaboration strength.

[0089] Furthermore, as an optional embodiment of the present invention, the step of performing correlation analysis on the standard interaction features to obtain key collaborative features includes:

[0090] The standard interaction features are processed into feature labels to obtain interaction feature labels;

[0091] Calculate the association strength value between the interactive feature labels;

[0092] Based on the association strength value, the collaborative interaction features in the standard interaction features are extracted;

[0093] Calculate the feature redundancy corresponding to the collaborative interaction features;

[0094] Based on the aforementioned feature redundancy, the collaborative interaction features are filtered to obtain key collaborative features.

[0095] The interactive feature labels are type identifiers assigned to each standard interactive feature through cluster analysis; the association strength value is a quantitative indicator that measures the closeness of the collaborative relationship between different feature labels; the collaborative interactive feature is the original feature set corresponding to the feature labels with high association strength values; the feature redundancy is a measure of the degree of information overlap between collaborative interactive features; and the key collaborative feature is a combination of core features with independent information value that is retained after redundancy screening.

[0096] Furthermore, the standard interaction features can be labeled using the K-means clustering method, dividing them into different feature categories based on their numerical distribution characteristics, and assigning a corresponding type label to each feature to obtain interaction feature labels. The Spearman rank correlation coefficient method can be used to calculate the association strength between the interaction feature labels, quantifying their association degree by analyzing the co-occurrence patterns of different feature labels in the samples. Highly correlated feature label groups can be extracted from the association strength values ​​based on a preset association strength threshold, thereby finding the original feature data corresponding to these labels to obtain collaborative interaction features. The feature redundancy corresponding to the collaborative interaction features can be calculated using the mutual information calculation method, measuring the amount of redundant information contained between any two features and establishing a feature redundancy matrix. The collaborative interaction features can be screened based on the maximum correlation and minimum redundancy criterion, minimizing the redundancy between features while ensuring the correlation between features and the collaborative operation index, and obtaining key collaborative features that are both representative and non-repetitive through iterative selection.

[0097] Furthermore, as an optional embodiment of the present invention, the step of calculating the cooperative operation index corresponding to the agent node by combining the node cooperative relationship graph and the key cooperative features includes:

[0098] The node collaboration graph is subjected to graph embedding processing to obtain the node structure feature vector;

[0099] The key collaborative features are reconstructed to obtain node behavior feature vectors;

[0100] Calculate the structural similarity between the node structural feature vectors, and determine the structural coordination index corresponding to the agent node based on the structural similarity;

[0101] Calculate the behavior matching degree between the node behavior feature vectors, and determine the behavior coordination index corresponding to the agent node based on the behavior matching degree;

[0102] By combining the structural coordination index and the behavioral coordination index, the coordinated operation index corresponding to the agent node is determined.

[0103] Wherein, the node structural feature vector is the positional feature representation of the node in the network topology extracted by a graph neural network; the node behavioral feature vector is a compact representation of key collaborative features reconstructed by an autoencoder; the structural similarity is a metric that quantifies the degree of similarity in the structural positions of nodes in the network; the behavioral matching degree is a correlation index that measures the consistency of behavioral features among nodes; the structural coordination index is an evaluation value reflecting the collaborative potential of nodes in the network structure, calculated based on structural similarity; and the behavioral coordination index is an evaluation value reflecting the collaborative performance of nodes in actual operation, calculated based on behavioral matching degree.

[0104] Furthermore, a graph attention network can be used to perform graph embedding processing on the node collaboration graph to learn the connection strength between nodes and their neighboring nodes, generating node structural feature vectors containing network topology information. A variational autoencoder can be used to reconstruct the key collaboration features, extracting the essential patterns of the features through the encoder and ensuring reconstruction accuracy through the decoder, resulting in dimensionality-reduced and noise-reduced node behavior feature vectors. The structural similarity between the node structural feature vectors can be calculated using the cosine similarity method, evaluating the proximity of nodes in the network structure based on the angle between them in the vector space. After normalization, the structural collaboration index corresponding to the agent node is determined, such as the structural features of a certain automotive parts production node with three vehicle assembly nodes and two raw material supply nodes. The mean cosine similarity of the vectors is 0.75. Min-Max normalization maps this mean to the [0,1] interval, resulting in a normalized value of 0.82. This result is the structural coordination index of the production node. The behavioral matching degree between the node behavioral feature vectors can be calculated using the Pearson correlation coefficient method to measure the linear correlation of the changing trends of different node behavioral features. After normalization, the behavioral coordination index corresponding to the agent node is determined, following the same principle as the structural coordination index calculation. A linear weighted fusion method can be used to combine the structural coordination index and the behavioral coordination index, assigning different weight ratios to them according to actual application requirements. A weighted summation is then used to calculate a comprehensive collaborative operation index that reflects the node's collaborative capabilities.

[0105] S2. Obtain historical interaction records of the industry chain network, extract the correlation features in the historical interaction records, and analyze the efficiency of advantage transfer between the intelligent agent nodes.

[0106] This invention, by acquiring historical interaction records of an industry chain network and extracting correlation features, can understand the connection strength and resource flow patterns of intelligent agent nodes in historical interactions, providing a basis for analyzing the efficiency of advantage transfer between nodes. The historical interaction records are log data of past transactions, collaborations, or information exchanges between intelligent agent nodes in the industry chain network, reflecting the interaction history and relationship accumulation between nodes. The correlation features are quantitative indicators representing the relationship characteristics between nodes extracted from the historical interaction records. Furthermore, the historical interaction records corresponding to the industry chain network can be obtained through an industry collaboration information database or a record of past transactions.

[0107] As an embodiment of the present invention, the extraction of associated features from the historical interaction records includes:

[0108] Identify the interaction record text in the historical interaction records and calculate the text information entropy corresponding to the interaction record text;

[0109] Based on the text information entropy, extract the core interaction records from the historical interaction records;

[0110] The core interaction records are processed in a structured manner to obtain field-structured interaction records;

[0111] Extract the interaction association factors from the field structure interaction records, perform association pattern recognition on the interaction association factors, and obtain association pattern features;

[0112] The associated pattern features are subjected to feature aggregation processing to obtain associated features.

[0113] Wherein, the interaction record text is the text characters in the historical interaction record; the text information entropy is a quantitative index obtained by calculating the information complexity of the interaction record text, used to evaluate the information richness of the interaction content; the core interaction record is key interaction data with information density higher than a preset threshold selected based on text information entropy; the field structure interaction record is a standardized data form after the core interaction record is structured and organized according to a predetermined field specification; the interaction association factor is a set of key elements representing the association characteristics between nodes extracted from the structured record; the association pattern feature is a set of features representing the association rules between nodes obtained by pattern recognition of the interaction association factor; the feature aggregation processing is a process of integrating multiple association pattern features into a unified feature vector through a weighted fusion algorithm; the association feature is a comprehensive feature vector representing the stable association relationship between agent nodes obtained after feature aggregation processing.

[0114] Furthermore, OCR recognition technology can be used to identify the interaction record text in the historical interaction records; the textual information entropy of the interaction record text can be calculated using the Shannon entropy formula combined with the TF-IDF algorithm to obtain a quantitative evaluation value; the core interaction records can be extracted based on textual information entropy using a dynamic threshold method, for example, retaining the top 25% of high-value interaction data in terms of information entropy; the core interaction records can be structured using predefined field templates combined with natural language processing technology to obtain field-structured interaction records containing standard fields such as interaction subject, interaction type, and interaction intensity; interaction correlation factors can be extracted from the field-structured interaction records using factor analysis to identify key variables affecting node association; and cluster analysis combined with graph pattern mining algorithms can be used. The interaction correlation factors are subjected to correlation pattern recognition to obtain correlation pattern features including temporal correlation patterns, strength correlation patterns, and stability correlation patterns. The weight coefficients of each correlation pattern feature can be determined by the analytic hierarchy process (AHP), and then the correlation pattern features are aggregated using a weighted average algorithm to obtain correlation features. For example, among the three identified correlation patterns, a weight of 0.4 is assigned to the temporal correlation pattern, a weight of 0.35 is assigned to the strength correlation pattern, and a weight of 0.25 is assigned to the stability correlation pattern. The comprehensive correlation feature value is obtained by weighted calculation. When the value is higher than 0.8, it is defined as an efficient correlation feature; between 0.6 and 0.8, it is defined as a medium-efficiency correlation feature; and below 0.6, it is defined as an inefficient correlation feature. Thus, a correlation feature that can comprehensively characterize the correlation characteristics between nodes is constructed.

[0115] This invention extracts correlation features from historical interaction records to analyze the efficiency of advantage transfer between intelligent agent nodes. This allows for an understanding of the resource flow efficiency and collaboration potential between nodes in the industry chain network, thus providing data support for subsequent optimization decisions. The advantage transfer efficiency is a comprehensive description of the speed, stability, and effectiveness of resource, information, or technology transfer between intelligent agent nodes.

[0116] As an embodiment of the present invention, the step of analyzing the advantage transfer efficiency between the agent nodes based on the association features includes:

[0117] The efficiency-related features corresponding to the agent nodes are filtered out from the associated features, and the real-time interaction data of the agent nodes are obtained.

[0118] Calculate the transmission efficiency index corresponding to the efficiency-related characteristics;

[0119] Based on the real-time interaction data, it is determined whether there is an interaction interruption between the intelligent agent nodes. If there is no interaction interruption, the superior transmission efficiency between the intelligent agent nodes is analyzed by combining the transmission efficiency index and the efficiency-related characteristics.

[0120] If an interaction interruption occurs, the interruption information related to the interaction interruption is extracted from the real-time interaction data, and the interruption impact level corresponding to the interaction interruption is analyzed based on the interruption information.

[0121] By combining the interruption impact level and the transmission efficiency index, the superior transmission efficiency between the agent nodes is analyzed.

[0122] The efficiency-related features are a subset of features directly related to the efficiency of advantage transfer, such as interaction frequency, resource matching degree, and transfer delay. The real-time interaction data refers to the latest data set collected by the agent node that reflects its immediate interaction status, including but not limited to recent transaction volume, message response time, resource availability, and cooperation status updates. The transfer efficiency index is a comprehensive indicator used to quantify the degree of influence of efficiency-related features on advantage transfer efficiency. The interruption information is specific descriptive data of the interaction interruption that has occurred in the real-time interaction data, recording in detail the time of the interruption, the identifier of the node involved, the interruption symptoms, the preliminary cause judgment, and the recovery measures taken. The interruption impact level is the level of damage caused to advantage transfer by the interaction interruption based on the interruption information.

[0123] Furthermore, efficiency-related features corresponding to the agent node can be selected from the associated features by using a feature screening model combined with domain knowledge rules; real-time interaction data of the agent node can be obtained through the aforementioned supply chain management system database or real-time data stream interface; based on the real-time interaction data, an interruption detection threshold is set and an anomaly detection algorithm is used to determine whether an interaction interruption exists; if an interaction interruption exists, interruption information can be extracted using event analysis and impact assessment technology, and a hierarchical evaluation model can be used to analyze the interruption impact level; finally, an efficiency fusion algorithm is used to combine the interruption impact level and the transmission efficiency index to complete a comprehensive analysis of the superior transmission efficiency, such as an efficiency fusion logic based on "interruption impact level priority - transmission efficiency index scenario adaptation": if the interruption impact level is severe, it is directly determined as "low transmission efficiency"; if it is mild, it is further determined as "high-efficiency transmission" or "general transmission" based on whether the transmission efficiency index is higher than the benchmark threshold.

[0124] Furthermore, as an optional embodiment of the present invention, calculating the transmission efficiency index corresponding to the efficiency-related feature includes:

[0125] Extract multi-source collaborative features corresponding to efficiency-related features, and determine the collaborative efficiency value corresponding to the efficiency-related features based on the multi-source collaborative features;

[0126] Extract the dynamic evolution characteristics of the efficiency-related features, analyze the transmission contribution of the dynamic evolution characteristics in advantage assessment, and calculate the evolutionary stability corresponding to the efficiency-related features.

[0127] Based on the dynamic evolution characteristics, assign evolution weight coefficients corresponding to the efficiency-related characteristics;

[0128] The current assessment requirements for the core advantages of the industrial chain are obtained, and based on the assessment requirements, the transfer adjustment margin and noise tolerance threshold of the efficiency-related characteristics are calculated.

[0129] By combining the cooperative efficiency value, the evolution weight coefficient, the transmission adjustment margin, the evolution stability, and the noise tolerance threshold, the transmission efficiency index corresponding to the efficiency-related feature is calculated.

[0130] Among them, the multi-source synergy feature is a comprehensive indicator reflecting the synchronicity and coordination between different efficiency sources, such as technical efficiency, resource allocation efficiency, and innovation transformation efficiency; the synergy efficiency value is a numerical value used to quantify the overall synergy level of the multi-source synergy feature; the dynamic evolution feature is the changing trend and fluctuation pattern of the efficiency-related feature over time; the transmission contribution is used to measure the degree of influence of the dynamic evolution feature on the transmission effect of the industrial chain advantage; the evolution stability is an indicator used to quantify the stability of the dynamic change process of the efficiency-related feature; the evolution weight coefficient is the influence weight assigned to the dynamic evolution feature according to its importance; the transmission adjustment margin is the efficiency adjustment space reserved to cope with emergencies under the current judgment requirements; and the noise tolerance threshold is the maximum limit of data fluctuations or interference that the efficiency-related feature can withstand without affecting the judgment conclusion.

[0131] Furthermore, the multi-source collaborative features can be comprehensively evaluated using the TOPSIS method (Approximation of Ideal Solution Ranking Method) to calculate their proximity to the ideal collaborative state, thereby determining the collaborative efficiency value. The contribution of the dynamic evolutionary features in advantage assessment can be quantified by calculating the Hurst exponent or analyzing their Markov transition matrix, and the evolutionary stability can be calculated based on the variance or coefficient of variation of the time series. The evolutionary weight coefficients can be assigned to the efficiency-related features using the entropy weight method or a subjective-objective combination weighting method based on the contribution of the contribution. The transmission adjustment margin can be calculated based on the stringency and priority of the assessment requirements, combined with the quantiles of historical performance data, and the noise tolerance threshold can be set using the control limit principle in the statistical process control (SPC) method.

[0132] Furthermore, as another optional embodiment of the present invention, the calculation of the transmission efficiency index corresponding to the efficiency-related feature using the following formula, combining the cooperative efficiency value, the evolutionary weight coefficient, the transmission adjustment margin, the evolutionary stability, and the noise tolerance threshold, includes:

[0133]

[0134] Where A represents, This represents the collaborative efficiency value. Represents the evolution weighting coefficient. This indicates the allowance for adjustment. Indicates evolutionary stability. Indicates the noise tolerance threshold. () represents the Sigmoid function.

[0135] The principle behind this formula is: The harmonic mean approach integrates collaborative efficiency and evolutionary weights, placing greater emphasis on balance compared to the geometric mean. It is particularly suitable for scenarios where there are significant differences in the magnitude of features. By mapping the transfer adjustment margin to the (0,1) interval using the Sigmoid function, both the nonlinearity of the adjustment capability and numerical stability are preserved. An exponential saturation function is used to describe the contribution characteristics of evolutionary stability. When the stability far exceeds the noise tolerance threshold, it tends to saturate, which is consistent with cognitive laws.

[0136] The formula's effectiveness is as follows: In the application of core advantages in the industrial chain, this formula can effectively handle the magnitude difference between collaborative efficiency and evolutionary weights, enhance the sensitivity to low-weight features through harmonic averaging, and optimize the response characteristics to adjustment margin and evolutionary stability by using a combination of Sigmoid function and exponential saturation function, thus significantly improving the accuracy and robustness of the transmission efficiency index assessment.

[0137] S3. Query the current output level and target output requirements of the industrial chain network to calculate the output gap of the industrial chain network. Combine the collaborative operation index, the advantage transmission efficiency and the output gap to set the node contribution position of the intelligent agent node in the industrial chain.

[0138] This invention calculates the output gap of the industrial chain network by querying the current output level and target output requirements of the industrial chain network. It can accurately quantify the degree of deviation between the industrial chain network and the expected output target in actual operation, objectively reflect the achievement of production efficiency, and provide a key basis for subsequently setting the node contribution position of the intelligent agent node in the industrial chain.

[0139] The current output level refers to the time-varying output data of the industrial chain, actually acquired through the production data acquisition system, typically recorded at daily or weekly intervals, reflecting the actual production efficiency of the industrial chain under the current environment. The target output requirement refers to a benchmark output value pre-set based on market demand or strategic planning, serving as a reference for assessing the compliance of the industrial chain's operation. For example, the quarterly target output value for a manufacturing cluster might be 10,000 units. The output gap refers to the difference between the current output level and the target output requirement, used to measure the degree of compliance with production efficiency. Furthermore, the current output level and target output requirement can be extracted from the industrial chain operation data through a data query interface written in Java.

[0140] As an embodiment of the present invention, the step of querying the current output level and target output requirement of the supply chain network to calculate the output gap of the supply chain network includes:

[0141] Extract the effective output data from the current output level;

[0142] Identify key gap elements and their corresponding performance values ​​from the effective output data, and obtain the element calibration values ​​corresponding to the key gap elements.

[0143] Calculate the element sensitivity coefficients corresponding to the key gap elements;

[0144] Based on the sensitivity coefficient of the factor, the performance value of the factor, and the calibration value of the factor, the output gap of the industrial chain network is calculated.

[0145] The effective output data refers to the core dataset that truly reflects the actual output efficiency of the industrial chain after removing invalid and redundant data from the full amount of data related to the current output of the industrial chain network. The key gap elements refer to the core dimension factors that directly affect the industrial chain network to achieve the target output requirements, covering factors such as capacity supply, efficiency transformation, quality assurance, and collaborative connection. The element performance value is the actual quantitative result of the key gap elements at the current output level and is the core indicator reflecting the current operating status of the elements. The element calibration value is determined based on the strategic target output requirements, industry benchmark levels, and policy compliance standards of the industrial chain network. It is the target benchmark value that the key gap elements need to achieve and is the core reference for measuring whether the element performance meets the standards. The element sensitivity coefficient is an indicator that quantifies the degree of influence and correlation strength of the key gap elements on the output level of the industrial chain. The larger the coefficient, the more significant the change in the output gap is due to the fluctuation of the element. The output gap refers to the quantitative difference between the comprehensive performance corresponding to the current effective output of the industrial chain network and the target output requirements. It is the core result indicator reflecting the fit between the industrial chain output capacity and the target requirements.

[0146] Furthermore, effective output data can be extracted through multi-dimensional data filtering algorithms, including threshold filtering, association rule filtering, and core link matching. For example, abnormally low-output data caused by equipment failure during production and duplicate statistical data in the circulation process can be removed, while qualified output data from continuous production periods and delivery data of core products can be retained. An element importance assessment system can be constructed using the analytic hierarchy process (AHP) combined with the entropy weight method to identify key gap elements from effective output data. Statistical modeling of data can be used to obtain the corresponding element performance values. Simultaneously, by benchmarking against leading companies in the industry, breaking down the quantitative indicators of strategic goals in the industrial chain, and referencing compliance standards required by policies, the element calibration values ​​corresponding to key gap elements can be determined. For example, the chip yield rate calibration value in an electronics industry chain can be referenced to the average level of leading companies in the industry, while the core component delivery cycle calibration value is determined based on the breakdown of strategic goals. Element sensitivity coefficients can be calculated using the elasticity coefficient method or grey relational analysis model. Taking the elasticity coefficient method as an example... By constructing a regression model of the rate of change of factor performance values ​​and the rate of change of output level, the factor sensitivity coefficient is obtained. For example, the higher the sensitivity coefficient of chip yield rate, the more significant its change has on the overall output level of the industrial chain. The ranking of sensitivity coefficients directly reflects the priority of factors' impact on output gap. Finally, the output gap of the industrial chain network is calculated based on a weighted summation model. The formula is: Output gap = Σ[Factor sensitivity coefficient × |Factor performance value − Factor calibration value| / Factor calibration value], where |Factor performance value − Factor calibration value| / Factor calibration value is the relative gap of a single factor. After weighting by the sensitivity coefficient, the comprehensive output gap is obtained. The value range is non-negative. The closer the value is to zero, the better the current output level matches the target requirements. For example, in a new energy vehicle industrial chain, the battery energy density performance value does not meet the calibration value, and the motor assembly efficiency performance value is lower than the calibration value. By weighting the relative gap of each factor with the sensitivity coefficient, the comprehensive output gap is obtained, which intuitively reflects the quantitative gap between the current output of the industrial chain and the target.

[0147] Furthermore, as an optional embodiment of the present invention, calculating the element sensitivity coefficient corresponding to the key gap element includes:

[0148] Independent component separation is performed on the key gap elements to obtain the element eigenmodes;

[0149] Analyze the correlation mapping relationship between the intrinsic modes of the elements;

[0150] Based on the aforementioned association mapping relationship, construct the element action path corresponding to the key gap element;

[0151] Based on the action path of the aforementioned elements, the node action intensity corresponding to the key gap elements is calculated.

[0152] Based on the node effect strength, calculate the element sensitivity coefficient corresponding to the key gap element.

[0153] The term "element intrinsic mode" refers to the basic mode that reflects the core attributes and essential mechanism of a factor after separating key gap factors into independent components and removing redundant interference components and cross-coupling effects. The "association mapping relationship" refers to the interaction relationships between different element intrinsic modes, including positive promoting relationships, negative restrictive relationships, direct relationships, and indirect relationships, which can reveal the direction and degree of correlation between modes. The "element action path" is a dynamic transmission link constructed based on the association mapping relationship, clarifying the specific transmission path and link connection relationship of the element intrinsic mode from its initial effect to its final impact on output. The "node effect strength" refers to the contribution and weight of the element intrinsic mode as a transmission node in the element action path to the overall impact transmission process; the higher the node effect strength, the stronger the mode's transmission dominance in the path. The "element sensitivity coefficient" is a quantitative indicator of the impact degree and correlation strength of key gap factors on the output gap of the industrial chain, obtained by combining the node effect strength corresponding to each element intrinsic mode with the degree of correlation mapping relationship, which can intuitively reflect the sensitivity of factor fluctuations to changes in output gap.

[0154] Furthermore, key gap elements can be independently component-separated using methods such as wavelet decomposition to remove noise and non-core related components, extracting the element intrinsic modes. For example, from key gap elements related to capacity supply, core intrinsic modes such as equipment operating status mode and capacity load mode can be separated. The correlation mapping relationships between different element intrinsic modes can be identified using methods such as mutual information analysis, grey relational analysis, or Pearson correlation analysis, clarifying the direction of action and the degree of correlation between modes. For example, grey relational analysis reveals a strong positive correlation mapping relationship between the equipment operating status mode and the unit time output mode. Finally, topological network modeling methods can be used to transform the correlation mapping relationships into directed transmission links, constructing element action paths and identifying the starting nodes, intermediate transmission nodes, and terminal impact nodes in the path. For example, constructing a topological network modeling path for equipment operating status mode can be used to identify the key gap elements. The direct impact path from state mode to output per unit time mode to the overall output level of the industrial chain, and the indirect impact path from equipment operating state mode to resource input efficiency mode to output per unit time mode to the overall output level of the industrial chain, can be analyzed. The node influence strength can be calculated using the PageRank algorithm. By combining the transmission frequency and correlation of the mode in each impact path, the node influence strength corresponding to each element's intrinsic mode can be determined. For example, if an element's intrinsic mode is a key transmission node in multiple core impact paths and has a high degree of correlation, its node influence strength will be correspondingly higher. Based on the node influence strength, the element sensitivity coefficient corresponding to the key gap element can be calculated using a node influence aggregation algorithm. For example, by combining the core dominance and transmission efficiency of the node in each impact path, the influence strength of all related nodes can be aggregated to obtain the element sensitivity coefficient.

[0155] This invention, by combining the collaborative operation index, the advantage transfer efficiency, and the output gap, sets the node contribution ranking of the intelligent agent node in the industry chain. This quantifies the relative importance of each node in the value creation of the industry chain and provides a basis for subsequent configuration and enhancement processing of the industry chain network. The node contribution ranking is the order of the intelligent agent node's comprehensive value creation capability in the industry chain. Furthermore, by combining the collaborative operation index, the advantage transfer efficiency, and the output gap, a ranking algorithm integrating multi-dimensional indicators is used to set the node contribution ranking of the intelligent agent node in the industry chain. For example, the product of the collaborative operation index and the advantage transfer efficiency is used as the numerator, and the output gap is used as the denominator to calculate the contribution score. Then, the nodes are ranked in descending order of the score to determine their ranking, thus obtaining the node contribution ranking.

[0156] S4. Obtain the real-time supply capacity and reserve level of the core resource system on which the industrial chain network depends, so as to calculate the total support of the core resource system. Based on the total support, calculate the maximum carrying capacity of the industrial chain network at the current stage, so as to determine the available resource allocation corresponding to the industrial chain network.

[0157] This invention obtains the real-time supply capacity and reserve level of the core resource system upon which the industrial chain network depends, and calculates the overall support capacity of the core resource system. This allows for accurate understanding of the supply status of various key resources, preventing industrial chain disruptions due to insufficient resource supply, and providing a reliable basis for calculating the maximum carrying capacity of the industrial chain network at the current stage. The core resource system is a collection of key resources supporting the operation of the industrial chain, including core production factors such as raw materials, equipment, technology, and personnel. The real-time supply capacity refers to the quantity and quality of resources that the resource system can stably provide at the current moment. The reserve level is the scale of resources that have been reserved but not yet used in the resource system. The overall support capacity is the total amount of resources that can actually support the operation of the industrial chain at the current stage, after considering the continuous supply potential of the real-time supply capacity and the buffer capacity of the reserve level. Furthermore, the real-time supply capacity and reserve level of the core resource system upon which the industrial chain network depends can be obtained through multi-source resource dynamic sensing and data fusion technology. For example, by integrating production data from the resource supply side, scheduling data from the circulation side, and inventory data from the reserve side, the real-time supply status and reserve volume of core resources can be accurately captured.

[0158] As an embodiment of the present invention, the step of obtaining the real-time supply capacity and reserve level of the core resource system upon which the industrial chain network depends, in order to calculate the overall support capacity of the core resource system, includes:

[0159] The real-time supply capacity is stabilized to obtain the continuous supply level;

[0160] The reserve level is divided into time periods to obtain the stock distribution sequence;

[0161] The system tolerance is obtained by performing a collaborative matching analysis between the continuous supply level and the stock distribution sequence.

[0162] The system's tolerance is comprehensively calibrated to obtain the effective support capacity;

[0163] Based on the actual effective support capacity and the preset operating requirement cycle, the total support capacity of the core resource system is calculated.

[0164] The sustained supply level is a representation of the stable supply capacity of resources obtained after eliminating short-term fluctuation factors; the stock distribution sequence is a serialized dataset formed by segmenting and statistically analyzing resource reserves according to the time dimension; the system tolerance is an indicator representing the ability of resource supply and reserves to maintain stable system operation when coping with demand fluctuations; and the effective support capacity is the actual resource capacity that can be used to support the operation of the industrial chain after comprehensively considering efficiency losses and safety redundancy.

[0165] Furthermore, the real-time supply capacity can be stabilized using the Hodrick-Prescott filtering method to separate trend components from cyclical fluctuations and obtain a sustainable supply level. The reserve level can be divided into time periods using an equal-frequency binning method to ensure that each time period has the same statistical significance, forming a stock distribution sequence. A system dynamics model can be used to perform a collaborative matching analysis between the sustainable supply level and the stock distribution sequence, and the system tolerance can be obtained by simulating stress tests under supply-demand equilibrium conditions. The system tolerance can be comprehensively calibrated based on resource utilization efficiency curves and safety margin requirements, introducing attenuation factors and buffer coefficients to calculate the effective support capacity. Combining the effective support capacity with a preset operating demand cycle, the overall support volume of the core resource system can be calculated using a capacity-time integral algorithm to ensure that the resource needs of the industrial chain network are met within a specified time frame.

[0166] This invention calculates the maximum carrying capacity of the industrial chain network at the current stage based on the overall support level, thereby determining the corresponding available resource allocation ratio. This avoids resource waste due to improper allocation and prevents operational interruptions due to insufficient resource supply. The maximum carrying capacity is the highest operational scale that the industrial chain network can stably maintain under current resource conditions, defining the resource usage boundaries of each link in the industrial chain. The available resource allocation ratio is the proportion of various resources allowed to be used in the industrial chain network at the current stage, providing a clear allocation standard for resource allocation. Furthermore, based on the overall support level, the invention calculates the maximum carrying capacity of the industrial chain network in relation to the basic consumption system of the industrial chain. The maximum carrying capacity is calculated using the ratio of the numbers. The basic consumption coefficient of the industrial chain is the amount of resources consumed per unit time required for the industrial chain network to maintain basic operation under the baseline operating conditions. It is obtained by analyzing the lowest resource consumption level during the stable operation period in historical operating data and comprehensively considering seasonal fluctuations and business continuity requirements. Based on the maximum carrying capacity, the available resource allocation ratio is determined by matching the carrying capacity with the resource demand characteristics of each link. For example, by matching the maximum carrying capacity with the resource demand intensity and supply timeliness requirements of each link, a resource allocation scheme that prioritizes key production links and dynamically adjusts auxiliary links can be established to achieve optimal resource allocation.

[0167] S5. Combining the node contribution ranking and the available resource allocation, perform configuration enhancement processing on the industry chain network to obtain the enhancement result.

[0168] This invention combines the node contribution ranking with the available resource allocation ratio to perform configuration enhancement processing on the industrial chain network, thereby improving the accuracy and systematicness of industrial chain resource allocation and optimizing the overall efficiency of the industrial chain.

[0169] Furthermore, combining the node contribution ranking and the available resource allocation, a configuration enhancement process is performed on the industry chain network to obtain the enhancement result. The specific steps are as follows: First, analyze the core influence dimensions of the node contribution ranking to identify high-value core nodes that need to be prioritized and inefficient nodes that need to be optimized; then, dissect the adaptation shortcomings of the available resource allocation to clarify the imbalance points and adjustment directions of various resources in different links; then, combine the correlation characteristics of the two to explore the core contradiction of mismatch between node value and resource supply and the space for collaborative optimization; finally, formulate a resource-oriented scheduling scheme, node function adaptation strategy, and link collaboration improvement details to form a complete industry chain network configuration enhancement scheme; finally, based on this scheme, the configuration enhancement process of the industry chain network is performed to obtain the enhancement result. For a more intuitive understanding of the configuration enhancement process of the industry chain network, please refer to [link to relevant documentation]. Figure 2 The diagram shown is a schematic of the configuration enhancement process for the industrial chain network provided by this invention. It should be noted that in this invention, Figure 2 The flowchart presented is only for the processing flow of supply chain network configuration enhancement and is not limited to the specific implementation of this configuration enhancement method in different actual industry scenarios.

[0170] Compared to the problems described in the background art, this invention, by collecting node operation data of each intelligent agent node in the industrial chain network and calculating the corresponding collaborative operation index based on the node operation data, can grasp the collaborative performance of different nodes in the industrial chain, providing a data foundation for subsequent evaluation of the overall operational efficiency of the industrial chain. Furthermore, by acquiring historical interaction records of the industrial chain network and extracting correlation features, this invention can understand the connection strength and resource flow patterns of intelligent agent nodes in historical interactions, providing a basis for analyzing the efficiency of advantage transfer between nodes. Moreover, by querying the current output level and target output requirements of the industrial chain network, this invention can calculate the output gap of the industrial chain network, accurately quantifying the actual operation of the industrial chain network relative to the expected output target. The degree of deviation objectively reflects the achievement of production efficiency standards, providing a key basis for subsequently setting the node contribution ranking of the intelligent agent nodes in the industrial chain. Furthermore, by acquiring the real-time supply capacity and reserve level of the core resource system upon which the industrial chain network depends, this invention calculates the overall support capacity of the core resource system, accurately grasping the supply status of various key resources, avoiding industrial chain operation interruptions due to insufficient resource supply, and providing a reliable basis for subsequently calculating the maximum carrying capacity of the industrial chain network at the current stage. Finally, by combining the node contribution ranking and the available resource allocation ratio, this invention performs configuration enhancement processing on the industrial chain network, improving the accuracy and systematicness of industrial chain resource allocation, and achieving optimization and improvement of the overall efficiency of the industrial chain. Therefore, the method for judging and strengthening the core advantages of the industrial chain based on multi-agent collaboration provided by this invention can improve the efficiency of judging and strengthening the core advantages of the industrial chain based on multi-agent collaboration.

[0171] like Figure 3 The diagram shown is a functional module diagram of a system for judging and strengthening the core advantages of the industrial chain based on multi-agent collaboration, according to the present invention.

[0172] The multi-agent collaborative supply chain core advantage assessment and enhancement system 200 described in this invention can be installed in an electronic device. Depending on the functions implemented, the multi-agent collaborative supply chain core advantage assessment and enhancement system may include a collaborative operation index calculation module 201, an advantage transmission efficiency analysis module 202, a node contribution ranking setting module 203, an available resource allocation ratio determination module 204, and a configuration enhancement processing module 205. The module described in this invention can also be called a unit, which refers to a series of computer program segments that can be executed by the processor of an electronic device and can perform a fixed function, stored in the memory of the electronic device.

[0173] In this embodiment of the invention, the functions of each module / unit are as follows:

[0174] The collaborative operation index calculation module 201 is used to collect node operation data of each intelligent agent node in the industrial chain network, and calculate the collaborative operation index corresponding to the intelligent agent node based on the node operation data.

[0175] The advantage transmission efficiency analysis module 202 is used to acquire historical interaction records of the industrial chain network, extract correlation features from the historical interaction records, and analyze the advantage transmission efficiency between the intelligent agent nodes.

[0176] The node contribution ranking setting module 203 is used to query the current output level and target output requirements of the industrial chain network, calculate the output gap of the industrial chain network, and set the node contribution ranking of the intelligent agent node in the industrial chain in combination with the collaborative operation index, the advantage transmission efficiency and the output gap.

[0177] The available resource allocation determination module 204 is used to obtain the real-time supply capacity and reserve level of the core resource system on which the industrial chain network depends, so as to calculate the total support of the core resource system, and based on the total support, calculate the maximum carrying capacity of the industrial chain network at the current stage, so as to determine the available resource allocation corresponding to the industrial chain network.

[0178] The configuration enhancement processing module 205 is used to combine the node contribution ranking and the available resource allocation to perform configuration enhancement processing on the industry chain network and obtain enhancement results.

[0179] In detail, the modules in the multi-agent collaborative supply chain core advantage assessment and enhancement system 200 described in this embodiment of the invention employ the same methods as described above. Figure 1 The method described herein is the same as the one for judging and strengthening the core advantages of the industrial chain based on multi-agent collaboration, and can produce the same technical effect, so it will not be elaborated here.

[0180] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0181] Finally, it should be noted that in the above embodiments, each embodiment can be combined with each other or independent. Deleting any one of them will not affect the technical implementation of other embodiments. The above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for assessing and strengthening core advantages in the industrial chain based on multi-agent collaboration, characterized in that, The method includes: Collect node operation data of each intelligent agent node in the industry chain network, and calculate the collaborative operation index corresponding to the intelligent agent node based on the node operation data. The calculation of the collaborative operation index corresponding to the intelligent agent node based on the node operation data includes: The node's operational data is subjected to quality screening to obtain valid operational data; Node interaction features are extracted from the valid operational data, and the node interaction features are normalized to obtain standard interaction features. Correlation analysis was performed on the standard interaction features to obtain key collaborative features; Based on the aforementioned key collaborative features, a node collaborative relationship graph is constructed; Based on the node collaboration relationship diagram and the key collaboration features, calculate the collaboration operation index corresponding to the agent node; Obtain historical interaction records of the industry chain network, extract correlation features from the historical interaction records, and analyze the advantage transfer efficiency between the intelligent agent nodes. The step of analyzing the advantage transfer efficiency between the intelligent agent nodes based on the correlation features includes: The efficiency-related features corresponding to the agent nodes are filtered out from the associated features, and the real-time interaction data of the agent nodes are obtained. Calculate the transmission efficiency index corresponding to the efficiency-related characteristics; Based on the real-time interaction data, it is determined whether there is an interaction interruption between the intelligent agent nodes. If there is no interaction interruption, the superior transmission efficiency between the intelligent agent nodes is analyzed by combining the transmission efficiency index and the efficiency-related characteristics. If an interaction interruption occurs, the interruption information related to the interaction interruption is extracted from the real-time interaction data, and the interruption impact level corresponding to the interaction interruption is analyzed based on the interruption information. By combining the interruption impact level and the transmission efficiency index, the superior transmission efficiency between the agent nodes is analyzed. The current output level and target output requirement of the industrial chain network are queried to calculate the output gap of the industrial chain network. Based on the collaborative operation index, the advantage transmission efficiency and the output gap, the node contribution position of the intelligent agent node in the industrial chain is set. The real-time supply capacity and reserve level of the core resource system on which the industrial chain network depends are obtained in order to calculate the total support of the core resource system. Based on the total support, the maximum carrying capacity of the industrial chain network at the current stage is calculated to determine the available resource allocation corresponding to the industrial chain network. By combining the node contribution ranking and the available resource allocation, a configuration enhancement process is performed on the industry chain network to obtain the enhancement result.

2. The method for judging and strengthening the core advantages of the industrial chain based on multi-agent collaboration as described in claim 1, characterized in that, The correlation analysis of the standard interaction features yields key collaborative features, including: The standard interaction features are processed into feature labels to obtain interaction feature labels; Calculate the association strength value between the interactive feature labels; Based on the association strength value, the collaborative interaction features in the standard interaction features are extracted; Calculate the feature redundancy corresponding to the collaborative interaction features; Based on the aforementioned feature redundancy, the collaborative interaction features are filtered to obtain key collaborative features.

3. The method for judging and strengthening the core advantages of the industrial chain based on multi-agent collaboration as described in claim 1, characterized in that, The calculation of the collaborative operation index corresponding to the agent node, combining the node collaborative relationship graph and the key collaborative features, includes: The node collaboration graph is subjected to graph embedding processing to obtain the node structure feature vector; The key collaborative features are reconstructed to obtain node behavior feature vectors; Calculate the structural similarity between the node structural feature vectors, and determine the structural coordination index corresponding to the agent node based on the structural similarity; Calculate the behavior matching degree between the node behavior feature vectors, and determine the behavior coordination index corresponding to the agent node based on the behavior matching degree; By combining the structural coordination index and the behavioral coordination index, the coordinated operation index corresponding to the agent node is determined.

4. The method for judging and strengthening the core advantages of the industrial chain based on multi-agent collaboration as described in claim 1, characterized in that, The extraction of association features from the historical interaction records includes: Identify the interaction record text in the historical interaction records and calculate the text information entropy corresponding to the interaction record text; Based on the text information entropy, extract the core interaction records from the historical interaction records; The core interaction records are processed in a structured manner to obtain field-structured interaction records; Extract the interaction association factors from the field structure interaction records, perform association pattern recognition on the interaction association factors, and obtain association pattern features; The associated pattern features are subjected to feature aggregation processing to obtain associated features.

5. The method for judging and strengthening the core advantages of the industrial chain based on multi-agent collaboration as described in claim 1, characterized in that, The calculation of the transmission efficiency index corresponding to the efficiency-related characteristics includes: Extract multi-source collaborative features corresponding to efficiency-related features, and determine the collaborative efficiency value corresponding to the efficiency-related features based on the multi-source collaborative features; Extract the dynamic evolution characteristics of the efficiency-related features, analyze the transmission contribution of the dynamic evolution characteristics in advantage assessment, and calculate the evolutionary stability corresponding to the efficiency-related features. Based on the dynamic evolution characteristics, assign evolution weight coefficients corresponding to the efficiency-related characteristics; The current assessment requirements for the core advantages of the industrial chain are obtained, and based on the assessment requirements, the transfer adjustment margin and noise tolerance threshold of the efficiency-related characteristics are calculated. By combining the cooperative efficiency value, the evolution weight coefficient, the transmission adjustment margin, the evolution stability, and the noise tolerance threshold, the transmission efficiency index corresponding to the efficiency-related feature is calculated.

6. The method for judging and strengthening the core advantages of the industrial chain based on multi-agent collaboration as described in claim 1, characterized in that, The process of querying the current output level and target output requirement of the industry chain network to calculate the output gap of the industry chain network includes: Extract the effective output data from the current output level; Identify key gap elements and their corresponding performance values ​​from the effective output data, and obtain the element calibration values ​​corresponding to the key gap elements. Calculate the element sensitivity coefficients corresponding to the key gap elements; Based on the sensitivity coefficient of the factor, the performance value of the factor, and the calibration value of the factor, the output gap of the industrial chain network is calculated.

7. The method for judging and strengthening the core advantages of the industrial chain based on multi-agent collaboration as described in claim 6, characterized in that, The calculation of the element sensitivity coefficient corresponding to the key gap element includes: Independent component separation is performed on the key gap elements to obtain the element eigenmodes; Analyze the correlation mapping relationship between the intrinsic modes of the elements; Based on the aforementioned association mapping relationship, construct the element action path corresponding to the key gap element; Based on the action path of the aforementioned elements, the node action intensity corresponding to the key gap elements is calculated. Based on the node effect strength, calculate the element sensitivity coefficient corresponding to the key gap element.

8. The method for judging and strengthening the core advantages of the industrial chain based on multi-agent collaboration as described in claim 7, characterized in that, The process of obtaining the real-time supply capacity and reserve level of the core resource system upon which the industrial chain network depends, in order to calculate the overall support capacity of the core resource system, includes: The real-time supply capacity is stabilized to obtain the continuous supply level; The reserve level is divided into time periods to obtain the stock distribution sequence; The system tolerance is obtained by performing a collaborative matching analysis between the continuous supply level and the stock distribution sequence. The system's tolerance is comprehensively calibrated to obtain the effective support capacity; Based on the actual effective support capacity and the preset operating requirement cycle, the total support capacity of the core resource system is calculated.

Images