Chain network self-healing control method and device

Abstract

The invention discloses a chain network self-healing control method and device, and relates to the field of network monitoring control, and the method comprises the steps: obtaining a first fault prediction result based on a preset fault diagnosis model according to the first transmission data of a current chain network; when the first fault prediction result accords with a preset fault condition, obtaining a first self-healing scheme according to the first transmission data, so that the current chain network executes a first self-healing operation according to the first self-healing scheme; obtaining second transmission data after the current chain network completes the first self-healing operation for a preset time length, and obtaining a first evaluation result according to the second transmission data based on a preset self-healing evaluation model; and when the first evaluation result does not conform to a preset self-healing success condition, correcting the first self-healing scheme according to the second transmission data based on a large language model to obtain a second self-healing scheme, so that the current chain network executes a second self-healing operation according to the second self-healing scheme. According to the invention, the self-healing effect of the chain communication network can be improved.

Description

Technical Field

[0001] The present application relates to the field of network monitoring and control, and in particular to a chain-type network self-healing control method and device. Background Art

[0002] In modern chain communication networks, nodes simultaneously carry out multiple functions, including data transmission and data collection. When a node fails, the stability of the entire network is affected. Therefore, how to monitor node failures and select corresponding self-healing strategies to control node self-healing is a key issue in the operation of chain communication networks.

[0003] In existing self-healing control methods for chain communication networks, heartbeat detection or end-to-end testing are usually selected as monitoring methods. However, this method has difficulty in ensuring monitoring accuracy, resulting in certain irrationalities in self-healing control, which in turn leads to a partial loss of self-healing effect. At the same time, after detecting a node failure, existing methods often use a hard match of a single indicator to locate the node, which is unable to accurately distinguish different types of failures. This leads to irrational control during self-healing after fault identification, and the self-healing effect needs to be improved. Therefore, how to improve the self-healing effect of chain communication networks remains a problem that needs to be solved urgently in existing technologies. Summary of the Invention

[0004] The present application provides a chain network self-healing control method and device to solve the technical problem that the self-healing effect of the existing chain communication network self-healing method is insufficient.

[0005] According to a first aspect of the embodiments of the present application, a chain network self-healing control method is provided, comprising:

[0006] Obtaining a first fault prediction result based on first transmission data of the current chain network and a preset fault diagnosis model; wherein the first transmission data includes a service transmission log and a node control log;

[0007] When the first fault prediction result meets the preset fault condition, obtaining a first self-healing solution according to the first transmission data, so that the current chain network performs a first self-healing operation according to the first self-healing solution;

[0008] Obtaining second transmission data after the current chain network completes the first self-healing operation for a preset time period, and obtaining a first evaluation result based on the second transmission data based on a preset self-healing evaluation model; wherein the self-healing evaluation model is trained based on a random forest model;

[0009] When the first evaluation result does not meet the preset self-healing success condition, based on the large language model, the first self-healing plan is corrected according to the second transmission data to obtain a second self-healing plan, so that the current chain network performs a second self-healing operation according to the second self-healing plan.

[0010] The present application first obtains a first fault prediction result based on a fault diagnosis model according to the first transmission data of the current chain network, and can predict the fault result of the current chain network according to the business transmission log and the node control log of the first transmission data, combined with judgment, to improve the accuracy of the first fault prediction result; then, when the first fault prediction result meets the fault condition, a first self-healing plan is obtained and executed according to the first transmission data, and then the second transmission data after self-healing is performed is obtained, and a first evaluation result is obtained based on the self-healing evaluation model; and then, when the first evaluation result does not meet the self-healing success condition, a second self-healing plan is corrected and executed based on the large language model. By evaluating the first self-healing and performing a second self-healing when the evaluation does not meet the self-healing success condition, the problem of insufficient self-healing effect of a single self-healing can be avoided, thereby improving the self-healing effect of the existing chain communication network.

[0011] In certain embodiments of the present application, obtaining a first fault prediction result based on the first transmission data of the current chain network and a preset fault diagnosis model specifically includes:

[0012] Splitting the first transmission data according to a preset time window size to obtain multiple groups of time window data;

[0013] Performing time series feature analysis on the business transmission log of each set of time window data to obtain multiple sets of business fluctuation features, and inputting the multiple sets of business fluctuation features into a preset business fault diagnosis model to obtain business fault prediction results;

[0014] Performing time series feature analysis on the node control log of each set of time window data to obtain multiple sets of node control features, and inputting the multiple sets of node control features into a preset node fault diagnosis model to obtain node fault prediction results;

[0015] A first fault prediction result is obtained according to the service fault prediction result and the node fault prediction result.

[0016] This application first divides the first transmission data into multiple groups of time window data according to the preset time window size, and then performs time series feature analysis on the business transmission log and node control log of each group of time window data. By dividing the data into time windows and analyzing them separately, the corresponding business fluctuation characteristics and node control characteristics can be accurately obtained, and then combined with the fault diagnosis model to obtain business fault prediction results and node fault prediction results, and further obtain the first fault prediction result, which can combine the business and nodes to predict faults in the current chain network, thereby improving the accuracy of fault prediction.

[0017] In certain embodiments of the present application, obtaining a first self-healing solution according to the first transmission data, so that the current chain network performs a first self-healing operation according to the first self-healing solution, specifically includes:

[0018] Determine fault information based on the first transmission data and a decision tree model, and obtain a corresponding first self-healing solution based on the fault information; wherein the fault information includes a fault node and a fault type;

[0019] The first self-healing solution is sent to the current chain network, so that the current chain network performs a first self-healing operation according to the first self-healing solution.

[0020] The present application first determines the fault information according to the first transmission data and based on the decision tree model, and then matches and executes the first self-healing plan. Through the decision tree model, the fault information can be accurately determined, thereby improving the matching degree between the first self-healing plan obtained according to the fault information and the current chain network, thereby improving the execution effect of the first self-healing plan.

[0021] In certain embodiments of the present application, determining fault information based on the first transmission data and a decision tree model, and matching and obtaining a corresponding first self-healing solution based on the fault information, specifically includes:

[0022] Determining a faulty node based on a decision tree model according to the node control log of the first transmission data;

[0023] According to the faulty node, screening the service transmission log of the first transmission data to obtain node service data, and determining the fault type based on a decision tree model according to the node service data and the node control log;

[0024] According to the fault node and the fault type, a corresponding first self-healing solution is obtained by matching.

[0025] The present application first determines the fault node based on the node control log of the first transmission data and the decision tree model, and then filters the business transmission log of the first transmission data through the fault node to obtain the node business data, and determines the fault type based on the decision tree model. It can narrow the business data search range through the fault node and avoid incorrect judgment of the fault type. Then, the first self-healing plan is matched by combining the fault node and the fault type, which can improve the matching degree between the first self-healing plan obtained according to the fault information and the current chain network, thereby improving the execution effect of the subsequent first self-healing plan.

[0026] In certain embodiments of the present application, the method of modifying the first self-healing solution based on the large language model and the second transmission data to obtain a second self-healing solution, so that the current chain network performs a second self-healing operation according to the second self-healing solution, specifically includes:

[0027] Based on the preset self-healing analysis keywords, the first self-healing solution is input into a preset large language model to ask a question and obtain a first answer handle;

[0028] Based on the preset self-healing correction keyword, according to the second transmission data, a question is asked through the first answer handle to obtain a second self-healing solution;

[0029] The second self-healing solution is sent to the current chain network, so that the current chain network performs a second self-healing operation according to the second self-healing solution.

[0030] This application first asks questions to a preset large language model based on the first self-healing solution to obtain a first answer handle, and then asks questions based on the second transmission data to obtain and execute the second self-healing solution. Questions are asked through the large language model to correct the first self-healing solution. The large language model can improve the interpretability of the second self-healing solution and further improve the matching degree between the first self-healing solution and the current chain network, thereby improving the execution effect of the second self-healing solution.

[0031] According to a second aspect of the embodiments of the present application, there is provided a chain network self-healing control device, comprising a first fault prediction module, a first self-healing execution module, a first self-healing evaluation module, and a second self-healing execution module;

[0032] The first fault prediction module is configured to obtain a first fault prediction result based on first transmission data of the current chain network and a preset fault diagnosis model; wherein the first transmission data includes a service transmission log and a node control log;

[0033] The first self-healing execution module is configured to obtain a first self-healing solution based on the first transmission data when the first fault prediction result meets a preset fault condition, so as to cause the current chain network to perform a first self-healing operation according to the first self-healing solution;

[0034] The first self-healing evaluation module is configured to obtain second transmission data after the current chain network completes the first self-healing operation for a preset time period, and obtain a first evaluation result based on the second transmission data based on a preset self-healing evaluation model; wherein the self-healing evaluation model is trained based on a random forest model;

[0035] The second self-healing execution module is used to, when the first evaluation result does not meet the preset self-healing success condition, correct the first self-healing plan based on the large language model and the second transmission data to obtain a second self-healing plan, so that the current chain network performs a second self-healing operation according to the second self-healing plan.

[0036] In certain embodiments of the present application, the first fault prediction module includes a first data segmentation unit, a service fault prediction unit, a node fault prediction unit, and a first fault prediction unit;

[0037] The first data segmentation unit is configured to segment the first transmission data into a plurality of time window data groups according to a preset time window size;

[0038] The service fault prediction unit is used to perform time series feature analysis on the service transmission log of each set of time window data to obtain multiple sets of service fluctuation features, and input the multiple sets of service fluctuation features into a preset service fault diagnosis model to obtain a service fault prediction result;

[0039] The node fault prediction unit is used to perform time series feature analysis on the node control log of each set of time window data to obtain multiple sets of node control features, and input the multiple sets of node control features into a preset node fault diagnosis model to obtain a node fault prediction result;

[0040] The first fault prediction unit is configured to obtain a first fault prediction result based on the service fault prediction result and the node fault prediction result.

[0041] In certain embodiments of the present application, the first self-healing execution module includes a self-healing solution acquisition unit and a first self-healing execution unit;

[0042] The self-healing solution acquisition unit is configured to determine fault information based on the first transmission data and a decision tree model, and to match and obtain a corresponding first self-healing solution based on the fault information; wherein the fault information includes a fault node and a fault type;

[0043] The first self-healing execution unit is configured to send the first self-healing solution to the current chain network, so that the current chain network performs a first self-healing operation according to the first self-healing solution.

[0044] In certain embodiments of the present application, the self-healing solution acquisition unit includes a fault node determination subunit, a fault type determination subunit, and a self-healing solution acquisition subunit;

[0045] The fault node determination subunit is configured to determine the fault node based on a decision tree model according to the node control log of the first transmission data;

[0046] The fault type determination subunit is configured to screen the service transmission log of the first transmission data according to the faulty node to obtain node service data, and determine the fault type based on a decision tree model according to the node service data and the node control log;

[0047] The self-healing solution acquisition subunit is configured to match and obtain a corresponding first self-healing solution according to the faulty node and the fault type.

[0048] In certain embodiments of the present application, the second self-healing execution module includes a first questioning unit, a second questioning unit, and a second self-healing execution unit;

[0049] The first questioning unit is configured to input the first self-healing solution into a preset large language model based on preset self-healing analysis keywords to ask a question and obtain a first answer handle;

[0050] The second questioning unit is configured to ask a question through the first answer handle based on a preset self-healing correction keyword and the second transmission data to obtain a second self-healing solution;

[0051] The second self-healing execution unit is configured to send the second self-healing solution to the current chain network, so that the current chain network performs a second self-healing operation according to the second self-healing solution.

[0052] The present application first obtains a first fault prediction result based on a fault diagnosis model according to the first transmission data of the current chain network, and can predict the fault result of the current chain network according to the business transmission log and the node control log of the first transmission data, combined with judgment, to improve the accuracy of the first fault prediction result; then, when the first fault prediction result meets the fault condition, a first self-healing plan is obtained and executed according to the first transmission data, and then the second transmission data after self-healing is performed is obtained, and a first evaluation result is obtained based on the self-healing evaluation model; and then, when the first evaluation result does not meet the self-healing success condition, a second self-healing plan is corrected and executed based on the large language model. By evaluating the first self-healing and performing a second self-healing when the evaluation does not meet the self-healing success condition, the problem of insufficient self-healing effect of a single self-healing can be avoided, thereby improving the self-healing effect of the existing chain communication network. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1: A flow chart of a chain network self-healing control method shown in certain embodiments of the present application;

[0054] Figure 2 : A module structure diagram of a chain network self-healing control device shown in certain embodiments of the present application. DETAILED DESCRIPTION

[0055] The embodiments of the present application are described in detail below. Examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals represent the same or similar elements or elements with the same or similar functions from beginning to end. The embodiments described below in conjunction with the accompanying drawings are exemplary and are only used to explain some embodiments of the present application and should not be understood as limiting the embodiments of the present application. Based on the embodiments shown in this application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0056] In the description of this application, it should be understood that the terms "first" and "second" are used for descriptive purposes only and should not be understood to indicate or imply relative importance or implicitly indicate the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include at least one of such features. In the description of this application, unless otherwise clearly specified, "multiple" and "several" mean two or more.

[0057] Existing self-healing control methods for chained communication networks require process monitoring, but existing monitoring methods struggle to ensure accuracy, leading to irrational control during self-healing and a partial loss of self-healing effectiveness. After detecting a node failure, existing self-healing control methods often use a rigid match based on a single metric, failing to accurately distinguish between different types of failures. This leads to irrational control during self-healing after fault identification, also resulting in a partial loss of self-healing effectiveness. Therefore, improving the self-healing effectiveness of chained communication networks remains a pressing issue in existing technologies.

[0058] Based on the above technical background, please refer to Figure 1 The embodiment of the present application provides a chain network self-healing control method, including steps S101 to S104, each of which is as follows:

[0059] Step S101: obtaining a first fault prediction result based on first transmission data of the current chain network and a preset fault diagnosis model; wherein the first transmission data includes a service transmission log and a node control log.

[0060] In certain embodiments of the present application, obtaining a first fault prediction result based on the first transmission data of the current chain network and a preset fault diagnosis model specifically includes:

[0061] Splitting the first transmission data according to a preset time window size to obtain multiple groups of time window data;

[0062] Performing time series feature analysis on the business transmission log of each set of time window data to obtain multiple sets of business fluctuation features, and inputting the multiple sets of business fluctuation features into a preset business fault diagnosis model to obtain business fault prediction results;

[0063] Performing time series feature analysis on the node control log of each set of time window data to obtain multiple sets of node control features, and inputting the multiple sets of node control features into a preset node fault diagnosis model to obtain node fault prediction results;

[0064] A first fault prediction result is obtained according to the service fault prediction result and the node fault prediction result.

[0065] In certain embodiments of the present application, by performing time series feature analysis on the business transmission logs of each group of time window data, a variety of business fluctuation characteristics can be obtained, including but not limited to the business data window transmission rate, the average business data transmission delay, and the average business data packet loss rate.

[0066] In certain embodiments of the present application, by performing time series feature analysis on the node control log of each set of time window data, a variety of node control features can be obtained, including but not limited to the average flow of the node window, changes in node device status, changes in node load and node status information; wherein the node status information includes the location of the node, the preset control parameters of the node and the node status changes.

[0067] In certain embodiments of the present application, the fault diagnosis model, the service fault diagnosis model or the node fault diagnosis model are all binary classification models based on deep learning, and the implementation methods include but are not limited to convolutional neural network model CNN, recursive neural network model RNN, bidirectional encoder model BERT and its improved models. The preferred implementation method is the bidirectional encoder model BERT.

[0068] This application first divides the first transmission data into multiple groups of time window data according to the preset time window size, and then performs time series feature analysis on the business transmission log and node control log of each group of time window data. By dividing the data into time windows and analyzing them separately, the corresponding business fluctuation characteristics and node control characteristics can be accurately obtained, and then combined with the fault diagnosis model to obtain business fault prediction results and node fault prediction results, and further obtain the first fault prediction result, which can combine the business and nodes to predict faults in the current chain network, thereby improving the accuracy of fault prediction.

[0069] Step S102: When the first fault prediction result meets the preset fault condition, a first self-healing solution is obtained according to the first transmission data, so that the current chain network performs a first self-healing operation according to the first self-healing solution.

[0070] In certain embodiments of the present application, the preset fault conditions include, but are not limited to, channel transmission obstruction, network interruption, or node failure.

[0071] In certain embodiments of the present application, obtaining a first self-healing solution according to the first transmission data, so that the current chain network performs a first self-healing operation according to the first self-healing solution, specifically includes:

[0072] Determine fault information based on the first transmission data and a decision tree model, and obtain a corresponding first self-healing solution based on the fault information; wherein the fault information includes a fault node and a fault type;

[0073] The first self-healing solution is sent to the current chain network, so that the current chain network performs a first self-healing operation according to the first self-healing solution.

[0074] In certain embodiments of the present application, the implementation of the decision tree model for determining fault information, fault nodes and fault types includes but is not limited to a random forest model, a gradient boosting tree model, an XGBoost model and an AdaBoost model, and the preferred implementation is a random forest model.

[0075] The present application first determines the fault information according to the first transmission data and based on the decision tree model, and then matches and executes the first self-healing plan. Through the decision tree model, the fault information can be accurately determined, thereby improving the matching degree between the first self-healing plan obtained according to the fault information and the current chain network, thereby improving the execution effect of the first self-healing plan.

[0076] In certain embodiments of the present application, determining fault information based on the first transmission data and a decision tree model, and matching and obtaining a corresponding first self-healing solution based on the fault information, specifically includes:

[0077] Determining a faulty node based on a decision tree model according to the node control log of the first transmission data;

[0078] According to the faulty node, screening the service transmission log of the first transmission data to obtain node service data, and determining the fault type based on a decision tree model according to the node service data and the node control log;

[0079] According to the fault node and the fault type, a corresponding first self-healing solution is obtained by matching.

[0080] The present application first determines the fault node based on the node control log of the first transmission data and the decision tree model, and then filters the business transmission log of the first transmission data through the fault node to obtain the node business data, and determines the fault type based on the decision tree model. It can narrow the business data search range through the fault node and avoid incorrect judgment of the fault type. Then, the first self-healing plan is matched by combining the fault node and the fault type, which can improve the matching degree between the first self-healing plan obtained according to the fault information and the current chain network, thereby improving the execution effect of the subsequent first self-healing plan.

[0081] Step S103: Obtain second transmission data after the current chain network completes the first self-healing operation for a preset time, and obtain a first evaluation result based on the second transmission data based on a preset self-healing evaluation model; wherein the self-healing evaluation model is obtained based on random forest model training.

[0082] In certain embodiments of the present application, the preset time length is preferably 30 minutes.

[0083] Step S104: When the first evaluation result does not meet the preset self-healing success condition, based on the large language model, the first self-healing plan is corrected according to the second transmission data to obtain a second self-healing plan, so that the current chain network performs a second self-healing operation according to the second self-healing plan.

[0084] In certain embodiments of the present application, the preset self-healing success conditions include but are not limited to channel transmission recovery, successful data retransmission, network communication recovery, communication recovery due to channel switching, successful node restart or successful standby node startup and other self-healing success conditions.

[0085] In certain embodiments of the present application, the implementation of the large language model includes but is not limited to open source models fine-tuned by GPT, Wenxin Yiyan, iFlytek Spark, Tongyi Qianwen, Kimi or Lora.

[0086] In certain embodiments of the present application, the method of modifying the first self-healing solution based on the large language model and the second transmission data to obtain a second self-healing solution, so that the current chain network performs a second self-healing operation according to the second self-healing solution, specifically includes:

[0087] Based on the preset self-healing analysis keywords, the first self-healing solution is input into a preset large language model to ask a question and obtain a first answer handle;

[0088] Based on the preset self-healing correction keyword, according to the second transmission data, a question is asked through the first answer handle to obtain a second self-healing solution;

[0089] The second self-healing solution is sent to the current chain network, so that the current chain network performs a second self-healing operation according to the second self-healing solution.

[0090] This application first asks questions to a preset large language model based on the first self-healing solution to obtain a first answer handle, and then asks questions based on the second transmission data to obtain and execute the second self-healing solution. Questions are asked through the large language model to correct the first self-healing solution. The large language model can improve the interpretability of the second self-healing solution and further improve the matching degree between the first self-healing solution and the current chain network, thereby improving the execution effect of the second self-healing solution.

[0091] Compared with the existing technology, the present application first obtains a first fault prediction result based on the fault diagnosis model according to the first transmission data of the current chain network, and can predict the fault result of the current chain network according to the business transmission log and the node control log of the first transmission data, thereby improving the accuracy of the first fault prediction result; then, when the first fault prediction result meets the fault condition, the first self-healing plan is obtained and executed according to the first transmission data, and then the second transmission data after self-healing is obtained, and the first evaluation result is obtained based on the self-healing evaluation model, and then when the first evaluation result does not meet the self-healing success condition, the second self-healing plan is corrected and executed based on the large language model. By evaluating the first self-healing and performing a second self-healing when the evaluation does not meet the self-healing success condition, the problem of insufficient self-healing effect of single self-healing can be avoided, thereby improving the self-healing effect of the existing chain communication network.

[0092] Corresponding to the above method, see Figure 2 , the embodiment of the present application provides a chain network self-healing control device, including a first fault prediction module 210, a first self-healing execution module 220, a first self-healing evaluation module 230 and a second self-healing execution module 240;

[0093] The first fault prediction module 210 is configured to obtain a first fault prediction result based on first transmission data of the current chain network and a preset fault diagnosis model; wherein the first transmission data includes a service transmission log and a node control log;

[0094] The first self-healing execution module 220 is configured to obtain a first self-healing solution based on the first transmission data when the first fault prediction result meets a preset fault condition, so as to cause the current chain network to perform a first self-healing operation according to the first self-healing solution;

[0095] The first self-healing evaluation module 230 is configured to obtain second transmission data after the current chain network completes the first self-healing operation for a preset time period, and obtain a first evaluation result based on the second transmission data based on a preset self-healing evaluation model; wherein the self-healing evaluation model is trained based on a random forest model;

[0096] The second self-healing execution module 240 is used to correct the first self-healing plan based on the large language model and the second transmission data when the first evaluation result does not meet the preset self-healing success condition, to obtain a second self-healing plan, so that the current chain network performs a second self-healing operation according to the second self-healing plan.

[0097] In certain embodiments of the present application, the first fault prediction module 210 includes a first data segmentation unit, a service fault prediction unit, a node fault prediction unit, and a first fault prediction unit;

[0098] The first data segmentation unit is configured to segment the first transmission data into a plurality of time window data groups according to a preset time window size;

[0099] The service fault prediction unit is used to perform time series feature analysis on the service transmission log of each set of time window data to obtain multiple sets of service fluctuation features, and input the multiple sets of service fluctuation features into a preset service fault diagnosis model to obtain a service fault prediction result;

[0100] The node fault prediction unit is used to perform time series feature analysis on the node control log of each set of time window data to obtain multiple sets of node control features, and input the multiple sets of node control features into a preset node fault diagnosis model to obtain a node fault prediction result;

[0101] The first fault prediction unit is configured to obtain a first fault prediction result based on the service fault prediction result and the node fault prediction result.

[0102] In certain embodiments of the present application, the first self-healing execution module 220 includes a self-healing solution acquisition unit and a first self-healing execution unit;

[0103] The self-healing solution acquisition unit is configured to determine fault information based on the first transmission data and a decision tree model, and to match and obtain a corresponding first self-healing solution based on the fault information; wherein the fault information includes a fault node and a fault type;

[0104] The first self-healing execution unit is configured to send the first self-healing solution to the current chain network, so that the current chain network performs a first self-healing operation according to the first self-healing solution.

[0105] In certain embodiments of the present application, the self-healing solution acquisition unit includes a fault node determination subunit, a fault type determination subunit, and a self-healing solution acquisition subunit;

[0106] The fault node determination subunit is configured to determine the fault node based on a decision tree model according to the node control log of the first transmission data;

[0107] The fault type determination subunit is configured to screen the service transmission log of the first transmission data according to the faulty node to obtain node service data, and determine the fault type based on a decision tree model according to the node service data and the node control log;

[0108] The self-healing solution acquisition subunit is configured to match and obtain a corresponding first self-healing solution according to the faulty node and the fault type.

[0109] In certain embodiments of the present application, the second self-healing execution module 240 includes a first questioning unit, a second questioning unit, and a second self-healing execution unit;

[0110] The first questioning unit is configured to input the first self-healing solution into a preset large language model based on preset self-healing analysis keywords to ask a question and obtain a first answer handle;

[0111] The second questioning unit is configured to ask a question through the first answer handle based on a preset self-healing correction keyword and the second transmission data to obtain a second self-healing solution;

[0112] The second self-healing execution unit is configured to send the second self-healing solution to the current chain network, so that the current chain network performs a second self-healing operation according to the second self-healing solution.

[0113] The present application first obtains a first fault prediction result based on a fault diagnosis model according to the first transmission data of the current chain network, and can predict the fault result of the current chain network according to the business transmission log and the node control log of the first transmission data, combined with judgment, to improve the accuracy of the first fault prediction result; then, when the first fault prediction result meets the fault condition, a first self-healing plan is obtained and executed according to the first transmission data, and then the second transmission data after self-healing is performed is obtained, and a first evaluation result is obtained based on the self-healing evaluation model; and then, when the first evaluation result does not meet the self-healing success condition, a second self-healing plan is corrected and executed based on the large language model. By evaluating the first self-healing and performing a second self-healing when the evaluation does not meet the self-healing success condition, the problem of insufficient self-healing effect of a single self-healing can be avoided, thereby improving the self-healing effect of the existing chain communication network.

[0114] It should be understood that the device provided in the embodiments of the present application corresponds to the aforementioned method, and the chain network self-healing control device provided in the embodiments of the present application can implement the chain network self-healing control method provided in any embodiment of the present application.

[0115] Adaptively, the embodiments of the present application further provide a computer device and a computer-readable storage medium.

[0116] The computer device comprises: a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor;

[0117] Wherein, when the processor executes the computer program, a chain network self-healing control method of the present application is implemented.

[0118] The computer-readable storage medium stores a plurality of instructions, which are suitable for loading by a processor to execute a chain network self-healing control method of the present application.

[0119] The above description is a partial embodiment of the present application, which further describes the purpose, technical solutions, and beneficial effects of the present application in detail. It should be understood that the above description of the partial embodiment of the present application is not to be construed as limiting the present application. In particular, it is pointed out that for those skilled in the art, any changes, modifications, equivalent substitutions, and variations made within the spirit and principles of the present application should be included within the scope of protection of the present application.

Claims

1. A chain network self-healing control method, characterized in that: include: Obtaining a first fault prediction result based on first transmission data of the current chain network and a preset fault diagnosis model; wherein the first transmission data includes a service transmission log and a node control log; When the first fault prediction result meets the preset fault condition, obtaining a first self-healing solution according to the first transmission data, so that the current chain network performs a first self-healing operation according to the first self-healing solution; Obtaining second transmission data after the current chain network completes the first self-healing operation for a preset time period, and obtaining a first evaluation result based on the second transmission data based on a preset self-healing evaluation model; wherein the self-healing evaluation model is trained based on a random forest model; When the first evaluation result does not meet the preset self-healing success condition, based on the large language model, the first self-healing plan is corrected according to the second transmission data to obtain a second self-healing plan, so that the current chain network performs a second self-healing operation according to the second self-healing plan.

2. A chain network self-healing control method according to claim 1, characterized in that: The first fault prediction result is obtained based on the first transmission data of the current chain network and a preset fault diagnosis model, specifically including: Splitting the first transmission data according to a preset time window size to obtain multiple groups of time window data; Performing time series feature analysis on the business transmission log of each set of time window data to obtain multiple sets of business fluctuation features, and inputting the multiple sets of business fluctuation features into a preset business fault diagnosis model to obtain business fault prediction results; Performing time series feature analysis on the node control log of each set of time window data to obtain multiple sets of node control features, and inputting the multiple sets of node control features into a preset node fault diagnosis model to obtain node fault prediction results; A first fault prediction result is obtained according to the service fault prediction result and the node fault prediction result.

3. A chain network self-healing control method according to claim 1, characterized in that: Obtaining a first self-healing solution according to the first transmission data, so that the current chain network performs a first self-healing operation according to the first self-healing solution, specifically includes: Determine fault information based on the first transmission data and a decision tree model, and obtain a corresponding first self-healing solution based on the fault information; wherein the fault information includes a fault node and a fault type; The first self-healing solution is sent to the current chain network, so that the current chain network performs a first self-healing operation according to the first self-healing solution.

4. A chain network self-healing control method according to claim 3, characterized in that: The determining fault information based on the first transmission data and a decision tree model, and obtaining a corresponding first self-healing solution based on the fault information, specifically includes: Determining a faulty node based on a decision tree model according to the node control log of the first transmission data; According to the faulty node, screening the service transmission log of the first transmission data to obtain node service data, and determining the fault type based on a decision tree model according to the node service data and the node control log; According to the fault node and the fault type, a corresponding first self-healing solution is obtained by matching.

5. A chain network self-healing control method according to claim 1, characterized in that: The method of modifying the first self-healing solution based on the large language model and the second transmission data to obtain a second self-healing solution, so that the current chain network performs a second self-healing operation according to the second self-healing solution, specifically includes: Based on the preset self-healing analysis keywords, the first self-healing solution is input into a preset large language model to ask a question and obtain a first answer handle; Based on the preset self-healing correction keyword, according to the second transmission data, a question is asked through the first answer handle to obtain a second self-healing solution; The second self-healing solution is sent to the current chain network, so that the current chain network performs a second self-healing operation according to the second self-healing solution.

6. A chain network self-healing control device, characterized in that: It includes a first fault prediction module, a first self-healing execution module, a first self-healing evaluation module and a second self-healing execution module; The first fault prediction module is configured to obtain a first fault prediction result based on first transmission data of the current chain network and a preset fault diagnosis model; wherein the first transmission data includes a service transmission log and a node control log; The first self-healing execution module is configured to obtain a first self-healing solution based on the first transmission data when the first fault prediction result meets a preset fault condition, so as to cause the current chain network to perform a first self-healing operation according to the first self-healing solution; The first self-healing evaluation module is configured to obtain second transmission data after the current chain network completes the first self-healing operation for a preset time period, and obtain a first evaluation result based on the second transmission data based on a preset self-healing evaluation model; wherein the self-healing evaluation model is trained based on a random forest model; The second self-healing execution module is used to, when the first evaluation result does not meet the preset self-healing success condition, correct the first self-healing plan based on the large language model and the second transmission data to obtain a second self-healing plan, so that the current chain network performs a second self-healing operation according to the second self-healing plan.

7. A chain network self-healing control device according to claim 6, characterized in that: The first fault prediction module includes a first data segmentation unit, a service fault prediction unit, a node fault prediction unit and a first fault prediction unit; The first data segmentation unit is configured to segment the first transmission data into a plurality of time window data groups according to a preset time window size; The service fault prediction unit is used to perform time series feature analysis on the service transmission log of each set of time window data to obtain multiple sets of service fluctuation features, and input the multiple sets of service fluctuation features into a preset service fault diagnosis model to obtain a service fault prediction result; The node fault prediction unit is used to perform time series feature analysis on the node control log of each set of time window data to obtain multiple sets of node control features, and input the multiple sets of node control features into a preset node fault diagnosis model to obtain a node fault prediction result; The first fault prediction unit is configured to obtain a first fault prediction result based on the service fault prediction result and the node fault prediction result.

8. The chain network self-healing control device according to claim 6, characterized in that: The first self-healing execution module includes a self-healing solution acquisition unit and a first self-healing execution unit; The self-healing solution acquisition unit is configured to determine fault information based on the first transmission data and a decision tree model, and to match and obtain a corresponding first self-healing solution based on the fault information; wherein the fault information includes a fault node and a fault type; The first self-healing execution unit is configured to send the first self-healing solution to the current chain network, so that the current chain network performs a first self-healing operation according to the first self-healing solution.

9. The chain network self-healing control device according to claim 8, characterized in that: The self-healing solution acquisition unit includes a fault node determination subunit, a fault type determination subunit and a self-healing solution acquisition subunit; The fault node determination subunit is configured to determine the fault node based on a decision tree model according to the node control log of the first transmission data; The fault type determination subunit is configured to screen the service transmission log of the first transmission data according to the faulty node to obtain node service data, and determine the fault type based on a decision tree model according to the node service data and the node control log; The self-healing solution acquisition subunit is configured to match and obtain a corresponding first self-healing solution according to the faulty node and the fault type.

10. The chain network self-healing control device according to claim 6, characterized in that: The second self-healing execution module includes a first questioning unit, a second questioning unit and a second self-healing execution unit; The first questioning unit is configured to input the first self-healing solution into a preset large language model based on preset self-healing analysis keywords to ask a question and obtain a first answer handle; The second questioning unit is configured to ask a question through the first answer handle based on a preset self-healing correction keyword and the second transmission data to obtain a second self-healing solution; The second self-healing execution unit is configured to send the second self-healing solution to the current chain network, so that the current chain network performs a second self-healing operation according to the second self-healing solution.

Images