Industrial chain autonomous regulation and control method and system based on brain cognitive thinking system
By processing data and training models based on a brain-based cognitive thinking system, a self-regulating model for supply chain resilience is generated. This solves the problem of resilience regulation in complex, giant supply chain networks, enabling rapid response to external shocks and daily optimization, thereby enhancing the stability and resilience of the supply chain.
Patent Information
- Application Number
- CN202510651854.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Priority Date
- 2024-05-21
- Filing Date
- 2025-05-20
- Publication Date
- 2025-10-28
AI Technical Summary
Existing technologies are insufficient to effectively predict and regulate the resilience of complex and massive industrial chain networks. They are slow to respond to sudden external interference, and it is difficult to quantify the risk propagation path, leading to major disasters such as congestion and chain breaks in key network links.
By employing a brain-based cognitive thinking system approach, through data acquisition, preprocessing, model building, and training, an artificial intelligence model for autonomous regulation of supply chain resilience is generated. This model is then combined with a resilience quantification measurement library to identify vulnerabilities and implement counterfactual intervention and regulation, thereby optimizing the supply chain structure and responding to external shocks.
It enhances the stability and resilience of the industrial chain, enabling early identification of risks, rapid adjustment of production plans, optimization of the supply chain, and strengthening of the ability to withstand external shocks, achieving independent control and continuous optimization.
Smart Images
Figure CN120851576A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the intersection of big data analysis and artificial intelligence technologies, and in particular to a method and system for autonomous regulation of the industrial chain based on brain cognitive thinking system. Background Technology
[0002] With the widespread application of big data, industrial internet, and artificial intelligence technologies in the interconnection between upstream, midstream, and downstream enterprises in the industrial chain, a complex and giant network composed of hundreds of millions of element entities or subsystems has been formed from the manufacturing end to the consumption end. How to apply artificial intelligence technology to enhance the resilience of the industrial chain, enable autonomous regulation, and proactively prevent major disasters has become an important technical challenge for enhancing the security and controllability of the industrial chain.
[0003] Currently, methods for resilient regulation of complex giant chain networks mainly rely on innovations in artificial intelligence technologies such as deep learning and complex system modeling, focusing primarily on two aspects: 1) Deep learning-based network resilience prediction and regulation, which mainly uses deep learning methods to identify and predict and regulate topological features such as the degree distribution of key nodes, average shortest path length, clustering coefficient, density, and connectivity, as well as risk propagation paths; 2) Network topology robustness regulation based on complex network theory, which characterizes the community structure and relationship patterns in the industrial chain network, identifies cooperation opportunities, identifies key partners, optimizes resource allocation and collaborative management, and formulates risk management strategies and countermeasures, thereby improving the overall efficiency and resilience of the industrial chain network.
[0004] However, the industrial chain is a complex, giant chain network with multiple nested layers, belonging to the category of complex systems research. Such complex giant chain networks contain a large number of physical nodes (enterprises, raw materials, equipment, energy, production, talent, technology, and capital, etc.), and the dynamic interactions between nodes and the system and environment involve frequent evolution of complex interactions such as dependence, competition, and correlation. This results in complex network seepage phase transition physics phenomena such as nonlinear multi-level coupling, dependency oscillation, disorder, and uncertainty. As a result, the above-mentioned industrial chain resilience regulation methods are slow to respond to sudden external disturbances (such as raw material supply interruptions or sudden changes in market demand), making it difficult to quantify the resilience of the industrial chain network, unable to effectively predict risk propagation paths, and even causing major disasters such as congestion, chain breaks, and decoupling of key links in the entire network in severe cases, making rapid recovery difficult. Summary of the Invention
[0005] In view of this, the present invention provides a method and system for autonomous regulation of the industrial chain based on the brain cognitive thinking system, which improves the stability and resilience of the industrial chain when facing external disturbances, and ensures the safe, reliable and efficient operation of the industrial chain.
[0006] To achieve the above objectives, this invention provides a method for autonomous regulation of the industrial chain based on a brain-cognitive thinking system, comprising the following steps: S1. Data acquisition steps: Acquire relevant data of the industrial chain and preprocess the relevant data of the industrial chain. Preprocessing of relevant data of the industrial chain includes data cleaning, data transformation, feature extraction and data standardization. S2. Data processing steps: Using the thinking cell machine of the brain cognitive thinking system, the preprocessed industrial chain related data in S1 is processed to remove outliers, fill in missing data, and dynamically characterize the evolution of dependency relationships. Then, the relationship between industrial chain nodes in the industrial chain related data is represented by graph embedding technology, and the measurement standard of multi-source heterogeneous multimodal data in the industrial chain related data is unified to form industrial chain regulation training dataset and test dataset. S3. Model building and training steps: Input the supply chain regulation training dataset and test dataset from S2 into the brain cognitive thinking system to generate an artificial intelligence model for autonomous regulation of supply chain resilience. Based on the empirical dataset, test the reliability, recall and accuracy of the artificial intelligence model for autonomous regulation of supply chain resilience to obtain the final autonomous regulation artificial intelligence model. S4. Autonomous Regulation Steps: Using the cognitive reasoning functional domain of the brain cognitive thinking system, based on the existing case library of industrial chain resilience regulation and empirical test data of industrial chain resilience, establish a quantitative measurement library of industrial chain network resilience, and through the quantitative measurement library of resilience and the final autonomous regulation artificial intelligence model in S3, identify industrial chain vulnerability and conduct counterfactual intervention regulation.
[0007] Optionally, obtain industry chain related data, including: Multiple data acquisition technologies were employed to obtain supply chain resilience datasets and empirical datasets; The various data acquisition technologies include: Internet of Things sensors, WebService, XML data exchange, and deep convolutional neural multimedia perception. The supply chain resilience dataset includes: enterprise data, supplier data, raw material supplier data, equipment provider data, energy supplier data, and manufacturer data; The empirical dataset includes: supply and demand matching case data, delivery case data, production readiness case data, inventory saturation case data, and rapid response to upstream and downstream demand fluctuations case data.
[0008] Optionally, the supply chain regulation training dataset and test dataset in S2 are input into the brain cognitive thinking system to generate an artificial intelligence model for autonomous regulation of supply chain resilience, including: From the training and test datasets of the industrial chain regulation, a dynamic training set and a dynamic test set are divided based on a sliding window. Adversarial training is introduced to generate robust samples. Through the brain cognitive thinking system, a universal model of industrial chain network permeation and a universal model of industrial chain network resilience dynamics are generated.
[0009] Alternatively, the solution process for the universal model of industry chain network seepage is as follows: in, This represents a universal model of industry chain network penetration. This represents the joint probability distribution of the topological clustering coefficients of k industry chain network nodes. This represents the degree of connectivity in the industry chain network. The coupling dynamic potential function representing the industrial chain network. Indicating the first in the industrial chain network Layer nodes and the first Dependencies between layer nodes An optimized calculation method for estimating critical points representing the degree connectivity of an industry chain network. This indicates the number of connection layers through dependency relationships.
[0010] Alternatively, the general model of industrial chain network resilience dynamics can be used, and the specific solution process is as follows: in, To describe the activity nodes at time t in the industrial chain status Evolutionary dynamic potential function Reflecting the first in the industrial chain Each feature node and its neighboring feature nodes The dynamic rules of interaction between them, where N represents the number of element nodes in the industrial chain.
[0011] Optionally, the resilience quantification measurement library of the industrial chain network is specifically used for: measuring the proportion of connected nodes in the industrial chain, determining the frequency of occurrence of nodes in the shortest path of the industrial chain network, identifying key nodes in the industrial chain network, solving for the maximum connected subgraph size, inventory turnover rate and safety stock coverage rate of the industrial chain network, and conducting risk propagation path analysis and stress test simulation of the industrial chain network.
[0012] Optionally, by using a resilience quantification measurement library and the final autonomous control AI model in S3, vulnerability identification and counterfactual intervention and control of the industrial chain can be carried out, including: if If the potential energy curve is stationary, then... The structure of online communities is healthy and stable, and no intervention or regulation is required; if If the potential energy curve is a periodic cosine or sine wave oscillation, then... The network community structure carries the risk of chain breakage; intervention and regulation can be implemented by increasing the proportion of nodes and adjacent edges in the industry chain network. If the potential energy curve is a chaotic oscillation, then... The network community structure suffers from severe congestion and lack of full connectivity. Intervention and regulation can be implemented by reducing the size of nodes in the industry chain network and pruning the proportion of connecting edges in the industry chain network.
[0013] This invention also provides an industry chain autonomous control system based on a brain-cognitive thinking system, and an industry chain autonomous control method based on a brain-cognitive thinking system applying any of the above claims, comprising a data acquisition module, a data processing module, a model building and training module, and an autonomous control module connected in sequence; wherein, The data acquisition module is used to acquire data related to the industrial chain and to preprocess the data, including data cleaning, data transformation, feature extraction and data standardization. The data processing module is used to employ the thinking cell machine of the brain cognitive thinking system to remove outliers, fill in missing data, and dynamically characterize the evolution of dependency relationships in the preprocessed industrial chain-related data in the data acquisition module. Then, the relationship between industrial chain nodes in the industrial chain-related data is represented by graph embedding technology, and the measurement standards of multi-source heterogeneous multimodal data in the industrial chain-related data are unified to form industrial chain regulation training dataset and test dataset. The model building and training module is used to input the supply chain regulation training dataset and test dataset from the data processing module into the brain cognitive thinking system to generate an artificial intelligence model for autonomous regulation of supply chain resilience. Based on the empirical dataset, the reliability, recall and accuracy of the artificial intelligence model for autonomous regulation of supply chain resilience are tested to obtain the final autonomous regulation artificial intelligence model. The autonomous regulation module is used to establish a quantitative measurement library of industrial chain network resilience based on the existing case library of industrial chain resilience regulation and empirical test data of industrial chain resilience, using the cognitive reasoning functional domain of the brain cognitive thinking system. Through the resilience quantitative measurement library and model building and training module, the final autonomous regulation artificial intelligence model is used to identify industrial chain vulnerability and intervene and regulate counterfactually.
[0014] As can be seen from the above technical solution, compared with the prior art, the present invention discloses a method and system for autonomous regulation of the industrial chain based on the brain cognitive thinking system, which has the following beneficial effects: (1) In terms of improving the robustness of the industrial chain, the system can predict and identify potential risks that may impact the industrial chain by analyzing historical and real-time data. For example, by monitoring the supply of key raw materials, changes in market demand, and natural or social events that may affect production, the system can respond in advance, adjust production plans, and ensure that the supply of key links is not affected.
[0015] (2) In terms of enhancing the resilience of the industrial chain, the system focuses on optimizing the industrial chain structure, strengthening weak links, and improving the overall ability to resist external shocks. For example, when faced with a global emergency such as the epidemic, the system can quickly adjust the supply chain network, find alternative suppliers, or accelerate the integration of the domestic industrial chain to reduce dependence on external markets.
[0016] (3) In terms of autonomous regulation, the system relies on advanced artificial intelligence technology, including machine learning, deep learning, large model and other technologies. Based on the brain cognitive thinking system, the resilience regulation of the industrial chain can be executed autonomously. This autonomy is not only reflected in the rapid response to emergencies, but also in the continuous optimization of daily operations, such as continuously optimizing the production process through intelligent algorithms to improve the efficiency of resource allocation. Attached Figure Description
[0017] To more clearly illustrate the technical solutions in the embodiments or related technologies of this application, the accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0018] Figure 1 A flowchart illustrating the supply chain autonomous control method based on a brain cognitive thinking system provided by this invention; Figure 2 This is a schematic diagram of the structure of the industrial chain autonomous control system based on the brain cognitive thinking system provided by the present invention. Detailed Implementation
[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0020] Figure 1 This is a flowchart illustrating the supply chain autonomous control method based on a brain-cognitive thinking system provided by the present invention. (Refer to...) Figure 1As shown, this invention provides a method for autonomous regulation of the industrial chain based on a brain-cognitive thinking system, comprising the following steps: S1. Data Acquisition Steps: Acquire relevant data from the industrial chain and preprocess the relevant data, including data cleaning, data transformation, feature extraction, and data standardization.
[0021] Optionally, industry chain-related data may include transaction records of upstream and downstream enterprises in the industry chain, market sentiment data, environmental monitoring data, emergency reports, supply chain distribution data, raw material supply data, and other data.
[0022] Optionally, obtain industry chain related data, including: Multiple data acquisition technologies were employed to obtain supply chain resilience datasets and empirical datasets.
[0023] Alternatively, various data acquisition technologies may include IoT sensors, Web Services, XML data exchange, and deep convolutional neural multimedia sensing technologies.
[0024] Optionally, the supply chain resilience dataset may include data such as enterprise data, supplier data, raw material supplier data, equipment provider data, energy supplier data, and manufacturer data.
[0025] Optionally, the empirical dataset may include case data such as supply and demand matching case data, delivery case data, production readiness case data, inventory saturation case data, and rapid response to upstream and downstream demand fluctuations case data.
[0026] S2. Data processing steps: Using the thinking cell machine of the brain cognitive thinking system, the preprocessed industrial chain related data in S1 is processed to remove outliers, fill in missing data, and dynamically characterize the evolution of dependency relationships. Then, the relationship between industrial chain nodes in the industrial chain related data is represented by graph embedding technology, and the measurement standard of multi-source heterogeneous multimodal data in the industrial chain related data is unified to form industrial chain regulation training dataset and test dataset.
[0027] Specifically, the thinking cell machine of the brain cognitive thinking system can be used to remove outliers, fill in missing data and dynamically characterize the evolution of dependency relationships in the preprocessed data obtained in S1. Then, graph embedding technology is used to characterize the relationship between nodes in the industrial chain in the industrial chain related data, and to ensure that the measurement standards of multi-source heterogeneous and multimodal data in the industrial chain related data are unified, so as to form the industrial chain regulation training dataset and test dataset.
[0028] S3. Model Building and Training Steps: Input the supply chain regulation training dataset and test dataset from S2 into the brain cognitive thinking system to generate an AI model for autonomous regulation of supply chain resilience. Based on the empirical dataset, test the reliability, recall, and accuracy of the AI model for autonomous regulation of supply chain resilience to obtain the final autonomous regulation AI model.
[0029] Specifically, in the model building and training steps, the supply chain regulation training dataset and test dataset in S2 can be input into the brain cognitive thinking system to train the supply chain resilience autonomous regulation artificial intelligence model, and the model is tested for reliability, recall and accuracy based on empirical data to finally form an autonomous regulation artificial intelligence model; Optionally, the supply chain regulation training dataset and test dataset in S2 are input into the brain cognitive thinking system to generate an artificial intelligence model for autonomous regulation of supply chain resilience, including: From the training and test datasets of the industrial chain regulation, a dynamic training set and a dynamic test set are divided based on a sliding window. Adversarial training is introduced to generate robust samples. Through the brain cognitive thinking system, a universal model of industrial chain network permeation and a universal model of industrial chain network resilience dynamics are generated.
[0030] Specifically, in S3, the autonomous regulation of artificial intelligence model construction and training involves using a brain cognitive thinking system to divide the supply chain regulation training dataset and test dataset into dynamic training set and dynamic test set based on a sliding window. This captures changes in data distribution in real time, avoids model performance degradation caused by sudden environmental changes, and introduces adversarial training to generate robust samples, thus generating a universal model of supply chain network permeation and a universal model of supply chain network resilience dynamics.
[0031] Optionally, the depth and width of the autonomous AI model's network structure can be adjusted in real time according to task requirements, activating lightweight sub-networks in resource-constrained scenarios and enabling full-parameter models for complex tasks.
[0032] S4. Autonomous Regulation Steps: Using the cognitive reasoning functional domain of the brain cognitive thinking system, based on the existing case library of industrial chain resilience regulation and empirical test data of industrial chain resilience, establish a quantitative measurement library of industrial chain network resilience, and through the quantitative measurement library of resilience and the final autonomous regulation artificial intelligence model in S3, identify industrial chain vulnerability and conduct counterfactual intervention regulation.
[0033] Specifically, in the autonomous regulation step, based on the existing case library of supply chain resilience regulation and empirical test data of supply chain resilience, and using the cognitive reasoning functional domain of the brain cognitive thinking system, a quantitative measurement library of supply chain network resilience can be established. The vulnerability of the supply chain can be identified through the final autonomous regulation artificial intelligence model in S3, and counterfactual intervention and regulation of the supply chain can be carried out in combination with diffusion dynamics.
[0034] Alternatively, the solution process for the universal model of industry chain network seepage is as follows: in, This represents a universal model of industry chain network penetration. express k The joint probability distribution of the phase transition of the topological clustering coefficient of each industrial chain network node. This represents the degree of connectivity in the industry chain network. The coupling dynamic potential function representing the industrial chain network. Indicating the first in the industrial chain network Layer nodes and the first Dependencies between layer nodes An optimized calculation method for estimating critical points representing the degree connectivity of an industry chain network. This indicates the number of connection layers through dependency relationships.
[0035] Alternatively, the general model of industrial chain network resilience dynamics can be used, and the specific solution process is as follows: in, To describe the industrial chain t Momentary activity nodes status Evolutionary dynamic potential function Reflecting the first in the industrial chain Each feature node and its neighboring feature nodes The dynamic rules of interaction between them, where N represents the number of element nodes in the industrial chain.
[0036] Optionally, the resilience quantification measurement library of the industrial chain network is specifically used for: measuring the proportion of connected nodes in the industrial chain, determining the frequency of occurrence of nodes in the shortest path of the industrial chain network, identifying key nodes in the industrial chain network, solving for the maximum connected subgraph size, inventory turnover rate and safety stock coverage rate of the industrial chain network, and conducting risk propagation path analysis and stress test simulation of the industrial chain network.
[0037] Specifically, in step S4, the autonomous control step, based on the universal model of the industrial chain network seepage, a quantitative measurement library of industrial chain network resilience is established from the existing case library of industrial chain resilience control and empirical test data of industrial chain resilience. This library can measure the proportion of connected nodes in the industrial chain, determine the frequency of occurrence of nodes in the shortest path, identify key nodes in the industrial chain network, solve for the maximum connected subgraph size, inventory turnover rate and safety stock coverage rate of the industrial chain network, and conduct risk propagation path analysis and stress test simulation of the industrial chain network.
[0038] Optionally, by using a resilience quantification measurement library and the final autonomous control AI model in S3, vulnerability identification and counterfactual intervention and control of the industrial chain can be carried out, including: if If the potential energy curve is stationary, then... The structure of online communities is healthy and stable, and no intervention or regulation is required; if If the potential energy curve is a periodic cosine or sine wave oscillation, then... The network community structure carries the risk of chain breakage; intervention and regulation can be implemented by increasing the proportion of nodes and adjacent edges in the industry chain network. If the potential energy curve is a chaotic oscillation, then... The network community structure suffers from severe congestion and lack of full connectivity. Intervention and regulation can be implemented by reducing the size of nodes in the industry chain network and pruning the proportion of connecting edges in the industry chain network.
[0039] Specifically, based on the resilience quantification measurement library and the final autonomous control artificial intelligence model, the specific content of supply chain vulnerability identification and counterfactual intervention control is as follows: First, it can be determined that in the universal model of supply chain network resilience dynamics... The potential energy curve type is used, and further combined with the case library of supply chain resilience regulation and empirical test data of supply chain resilience in the resilience quantification measurement library, to identify supply chain vulnerability and implement counterfactual intervention regulation.
[0040] if If the potential energy curve is stationary, then... The structure of online communities is healthy and stable, requiring no adjustment, i.e., no intervention or regulation is needed; if If the potential energy curve is a periodic cosine or sine wave oscillation, then... The risk of chain breakage exists in network community structures, which can be mitigated by increasing the proportion of nodes and adjacent edges in the industry chain network; if If the potential energy curve is a chaotic oscillation, then... The network community structure suffers from severe congestion and full connectivity issues, which can be resolved by reducing the node size in the industry chain network and pruning the proportion of connecting edges in the industry chain network.
[0041] Figure 2This is a schematic diagram of the structure of the industrial chain autonomous control system based on the brain cognitive thinking system provided by the present invention, with reference to... Figure 2 As shown, this invention provides an autonomous supply chain control system based on a brain-cognitive thinking system, and applies any of the above-mentioned methods for autonomous supply chain control based on a brain-cognitive thinking system, comprising a data acquisition module, a data processing module, a model building and training module, and an autonomous control module connected in sequence; wherein, The data acquisition module is used to acquire data related to the industrial chain and to preprocess the data, including data cleaning, data transformation, feature extraction and data standardization. The data processing module is used to employ the thinking cell machine of the brain cognitive thinking system to remove outliers, fill in missing data, and dynamically characterize the evolution of dependency relationships in the preprocessed industrial chain-related data in the data acquisition module. Then, the relationship between industrial chain nodes in the industrial chain-related data is represented by graph embedding technology, and the measurement standards of multi-source heterogeneous multimodal data in the industrial chain-related data are unified to form industrial chain regulation training dataset and test dataset. The model building and training module is used to input the supply chain regulation training dataset and test dataset from the data processing module into the brain cognitive thinking system to generate an artificial intelligence model for autonomous regulation of supply chain resilience. Based on the empirical dataset, the reliability, recall and accuracy of the artificial intelligence model for autonomous regulation of supply chain resilience are tested to obtain the final autonomous regulation artificial intelligence model. The autonomous regulation module is used to establish a quantitative measurement library of industrial chain network resilience based on the existing case library of industrial chain resilience regulation and empirical test data of industrial chain resilience, using the cognitive reasoning functional domain of the brain cognitive thinking system. Through the resilience quantitative measurement library and model building and training module, the final autonomous regulation artificial intelligence model is used to identify industrial chain vulnerability and intervene and regulate counterfactually.
[0042] Optionally, obtain industry chain related data, including: Multiple data acquisition technologies were employed to obtain supply chain resilience datasets and empirical datasets; The various data acquisition technologies include: Internet of Things sensors, WebService, XML data exchange, and deep convolutional neural multimedia perception. The supply chain resilience dataset includes: enterprise data, supplier data, raw material supplier data, equipment provider data, energy supplier data, and manufacturer data; The empirical dataset includes: supply and demand matching case data, delivery case data, production readiness case data, inventory saturation case data, and rapid response to upstream and downstream demand fluctuations case data.
[0043] Optionally, the supply chain regulation training dataset and test dataset from the data processing module are input into the brain cognitive thinking system to generate an artificial intelligence model for autonomous regulation of supply chain resilience, including: From the training and test datasets of the industrial chain regulation, a dynamic training set and a dynamic test set are divided based on a sliding window. Adversarial training is introduced to generate robust samples. Through the brain cognitive thinking system, a universal model of industrial chain network permeation and a universal model of industrial chain network resilience dynamics are generated.
[0044] Alternatively, the solution process for the universal model of industry chain network seepage is as follows: in, This represents a universal model of industry chain network penetration. express k The joint probability distribution of the phase transition of the topological clustering coefficient of each industrial chain network node. This represents the degree of connectivity in the industry chain network. The coupling dynamic potential function representing the industrial chain network. Indicating the first in the industrial chain network Layer nodes and the first Dependencies between layer nodes An optimized calculation method for estimating critical points representing the degree connectivity of an industry chain network. This indicates the number of connection layers through dependency relationships.
[0045] Alternatively, the general model of industrial chain network resilience dynamics can be used, and the specific solution process is as follows: in, To describe the industrial chain t Momentary activity nodes status Evolutionary dynamic potential function Reflecting the first in the industrial chain Each feature node and its neighboring feature nodes The dynamic rules of interaction between them, where N represents the number of element nodes in the industrial chain.
[0046] Optionally, the resilience quantification measurement library of the industrial chain network is specifically used for: measuring the proportion of connected nodes in the industrial chain, determining the frequency of occurrence of nodes in the shortest path of the industrial chain network, identifying key nodes in the industrial chain network, solving for the maximum connected subgraph size, inventory turnover rate and safety stock coverage rate of the industrial chain network, and conducting risk propagation path analysis and stress test simulation of the industrial chain network.
[0047] Optionally, by using a resilience quantification measurement library and the final autonomous control artificial intelligence model in the model building and training module, vulnerability identification and counterfactual intervention and control of the industrial chain can be carried out, including: if If the potential energy curve is stationary, then... The structure of online communities is healthy and stable, and no intervention or regulation is required; if If the potential energy curve is a periodic cosine or sine wave oscillation, then... The network community structure carries the risk of chain breakage; intervention and regulation can be implemented by increasing the proportion of nodes and adjacent edges in the industry chain network. If the potential energy curve is a chaotic oscillation, then... The network community structure suffers from severe congestion and lack of full connectivity. Intervention and regulation can be implemented by reducing the size of nodes in the industry chain network and pruning the proportion of connecting edges in the industry chain network.
[0048] As can be seen from the above technical solution, compared with the prior art, the present invention discloses a method and system for autonomous regulation of the industrial chain based on the brain cognitive thinking system, which has the following beneficial effects: (1) In terms of improving the robustness of the industrial chain, the system can predict and identify potential risks that may impact the industrial chain by analyzing historical and real-time data. For example, by monitoring the supply of key raw materials, changes in market demand, and natural or social events that may affect production, the system can respond in advance, adjust production plans, and ensure that the supply of key links is not affected.
[0049] (2) In terms of enhancing the resilience of the industrial chain, the system focuses on optimizing the industrial chain structure, strengthening weak links, and improving the overall ability to resist external shocks. For example, when faced with a global emergency such as the epidemic, the system can quickly adjust the supply chain network, find alternative suppliers, or accelerate the integration of the domestic industrial chain to reduce dependence on external markets.
[0050] (3) In terms of autonomous regulation, the system relies on advanced artificial intelligence technology, including machine learning, deep learning, large model and other technologies. Based on the brain cognitive thinking system, the resilience regulation of the industrial chain can be executed autonomously. This autonomy is not only reflected in the rapid response to emergencies, but also in the continuous optimization of daily operations, such as continuously optimizing the production process through intelligent algorithms to improve the efficiency of resource allocation.
[0051] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to the method section.
[0052] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims
1. A method for autonomous regulation of the industrial chain based on a brain-cognitive thinking system, characterized in that, Includes the following steps: S1. Data acquisition steps: Acquire relevant data of the industrial chain and preprocess the relevant data of the industrial chain. Preprocessing of relevant data of the industrial chain includes data cleaning, data transformation, feature extraction and data standardization. S2. Data processing steps: Using the thinking cell machine of the brain cognitive thinking system, the preprocessed industrial chain related data in S1 is processed to remove outliers, fill in missing data, and dynamically characterize the evolution of dependency relationships. Then, the relationship between industrial chain nodes in the industrial chain related data is represented by graph embedding technology, and the measurement standard of multi-source heterogeneous multimodal data in the industrial chain related data is unified to form industrial chain regulation training dataset and test dataset. S3. Model building and training steps: Input the supply chain regulation training dataset and test dataset from S2 into the brain cognitive thinking system to generate an artificial intelligence model for autonomous regulation of supply chain resilience. Based on the empirical dataset, test the reliability, recall and accuracy of the artificial intelligence model for autonomous regulation of supply chain resilience to obtain the final autonomous regulation artificial intelligence model. S4. Autonomous Regulation Steps: Using the cognitive reasoning functional domain of the brain cognitive thinking system, based on the existing case library of industrial chain resilience regulation and empirical test data of industrial chain resilience, establish a quantitative measurement library of industrial chain network resilience, and through the quantitative measurement library of resilience and the final autonomous regulation artificial intelligence model in S3, identify industrial chain vulnerability and conduct counterfactual intervention regulation.
2. The method for autonomous regulation of the industrial chain based on a brain-cognitive thinking system according to claim 1, characterized in that, Obtain relevant data from the industry chain, including: Multiple data acquisition technologies were employed to obtain supply chain resilience datasets and empirical datasets; The various data acquisition technologies include: Internet of Things sensors, WebService, XML data exchange, and deep convolutional neural multimedia perception. The supply chain resilience dataset includes: enterprise data, supplier data, raw material supplier data, equipment provider data, energy supplier data, and manufacturer data; The empirical dataset includes: supply and demand matching case data, delivery case data, production readiness case data, inventory saturation case data, and rapid response to upstream and downstream demand fluctuations case data.
3. The method for autonomous regulation of the industrial chain based on a brain-cognitive thinking system according to claim 1, characterized in that, The training and testing datasets for supply chain regulation in S2 are input into the brain-cognitive thinking system to generate an artificial intelligence model for autonomous regulation of supply chain resilience, including: From the training and test datasets of the industrial chain regulation, a dynamic training set and a dynamic test set are divided based on a sliding window. Adversarial training is introduced to generate robust samples. Through the brain cognitive thinking system, a universal model of industrial chain network permeation and a universal model of industrial chain network resilience dynamics are generated.
4. The method for autonomous regulation of the industrial chain based on a brain-cognitive thinking system according to claim 3, characterized in that, The general solution process for the supply chain network seepage model is as follows: in, This represents a universal model of industry chain network penetration. express k The joint probability distribution of the phase transition of the topological clustering coefficient of each industrial chain network node. This represents the degree of connectivity in the industry chain network. The coupling dynamic potential function representing the industrial chain network. Indicating the first in the industrial chain network Layer nodes and the first Dependencies between layer nodes An optimized calculation method for estimating critical points representing the degree connectivity of an industry chain network. This indicates the number of connection layers through dependency relationships.
5. The method for autonomous regulation of the industrial chain based on a brain-cognitive thinking system according to claim 3, characterized in that, The general solution process for the supply chain network resilience dynamics model is as follows: in, To describe the industrial chain t Momentary activity nodes status Evolutionary dynamic potential function Reflecting the first in the industrial chain Each feature node and its neighboring feature nodes The dynamic rules of interaction between them, where N represents the number of element nodes in the industrial chain.
6. The method for autonomous regulation of the industrial chain based on a brain-cognitive thinking system according to claim 1, characterized in that, The resilience quantification measurement library of the industrial chain network is specifically used for: measuring the proportion of connected nodes in the industrial chain, determining the frequency of nodes in the industrial chain network in the shortest path, identifying key nodes in the industrial chain network, solving for the maximum connected subgraph size, inventory turnover rate and safety stock coverage rate of the industrial chain network, and conducting risk propagation path analysis and stress test simulation of the industrial chain network.
7. The method for autonomous regulation of the industrial chain based on a brain-cognitive thinking system according to claim 5, characterized in that, Through a resilience quantification measurement library and the final autonomous control AI model in S3, vulnerability identification and counterfactual intervention in the industrial chain are conducted, including: if If the potential energy curve is stationary, then... The structure of online communities is healthy and stable, and no intervention or regulation is required; if If the potential energy curve is a periodic cosine or sine wave oscillation, then... The network community structure carries the risk of chain breakage; intervention and regulation can be implemented by increasing the proportion of nodes and adjacent edges in the industry chain network. If the potential energy curve is a chaotic oscillation, then... The network community structure suffers from severe congestion and lack of full connectivity. Intervention and regulation can be implemented by reducing the size of nodes in the industry chain network and pruning the proportion of connecting edges in the industry chain network.
8. A self-regulating supply chain system based on brain cognitive thinking system, characterized in that, A method for autonomous regulation of an industrial chain based on a brain-cognitive thinking system, as described in any one of claims 1-7, comprises, in sequence, a data acquisition module, a data processing module, a model building and training module, and an autonomous regulation module; wherein, The data acquisition module is used to acquire data related to the industrial chain and to preprocess the data, including data cleaning, data transformation, feature extraction and data standardization. The data processing module is used to employ the thinking cell machine of the brain cognitive thinking system to remove outliers, fill in missing data, and dynamically characterize the evolution of dependency relationships in the preprocessed industrial chain-related data in the data acquisition module. Then, the relationship between industrial chain nodes in the industrial chain-related data is represented by graph embedding technology, and the measurement standards of multi-source heterogeneous multimodal data in the industrial chain-related data are unified to form industrial chain regulation training dataset and test dataset. The model building and training module is used to input the supply chain regulation training dataset and test dataset from the data processing module into the brain cognitive thinking system to generate an artificial intelligence model for autonomous regulation of supply chain resilience. Based on the empirical dataset, the reliability, recall and accuracy of the artificial intelligence model for autonomous regulation of supply chain resilience are tested to obtain the final autonomous regulation artificial intelligence model. The autonomous regulation module is used to establish a quantitative measurement library of industrial chain network resilience based on the existing case library of industrial chain resilience regulation and empirical test data of industrial chain resilience, using the cognitive reasoning functional domain of the brain cognitive thinking system. Through the resilience quantitative measurement library and model building and training module, the final autonomous regulation artificial intelligence model is used to identify industrial chain vulnerability and intervene and regulate counterfactually.