Database fault diagnosis decision method, system and readable storage medium

Abstract

The present invention relates to the field of database technology, and in particular to a database fault diagnosis decision method, system and readable storage medium, the method comprising: building a decision center based on a large language model; using the decision center to receive fault information and parse it; based on the parsed information, the decision center builds a dynamic decision tree, the dynamic decision tree can dynamically adjust its branches and decision points according to known fault modes and real-time fault information; the decision center distributes tasks to a multi-intelligent agent network, the multi-intelligent agent network is configured with multiple intelligent agents, each intelligent agent focuses on a specific type of fault diagnosis task, each intelligent agent performs the assigned task and feeds back the result to the decision center; the decision center updates the dynamic decision tree according to the feedback result, and generates a fault diagnosis report. The present invention realizes high intelligence and automation of fault diagnosis, reduces the need for manual intervention, and improves the efficiency and accuracy of fault diagnosis.

Description

Technical Field

[0001] The present invention relates to the field of database technology, and in particular to a database fault diagnosis decision method, system and readable storage medium. Background Art

[0002] Database systems are the core infrastructure for data storage and management. With the explosive growth of data volume and the increasing complexity of business needs, efficient operation of databases and rapid fault diagnosis have become critical. Existing database fault diagnosis technologies mainly include the following:

[0003] 1) Rule-based fault diagnosis system: This type of system usually presets a series of fault detection rules and response strategies. When the system detects an abnormality, it triggers the corresponding fault handling process according to the predefined rules. However, this static rule-based method is difficult to cope with the ever-changing fault mode and complex system environment.

[0004] 2) Fault prediction based on machine learning: In recent years, machine learning technology has been applied to database fault prediction, which predicts potential future faults by analyzing historical data. Although this method has improved the accuracy of fault prediction to a certain extent, it still faces challenges in real-time and dynamic adaptability, especially in terms of immediate diagnosis and repair after a fault occurs.

[0005] 3) Expert system: Expert systems diagnose and solve problems by simulating the knowledge and decision-making process of human experts. However, the knowledge base of expert systems is updated slowly, and they are slow to respond to emerging fault types. They rely on the experience and intuition of experts and lack sufficient generalization ability and self-optimization mechanism.

[0006] The above-mentioned prior art mainly has the following defects:

[0007] Lack of dynamic adaptability: Traditional methods cannot quickly adapt to changes in database system architecture and sudden failures, resulting in delayed fault diagnosis and high misdiagnosis rates;

[0008] Over-reliance on manual intervention: Whether based on rules or expert systems, they are too dependent on manual intervention and cannot achieve fully automated diagnosis, increasing human resource costs;

[0009] Knowledge update lags: The update speed of rules and expert knowledge base cannot keep up with the development of technology and changes in failure modes, which reduces the timeliness and effectiveness of the diagnosis system;

[0010] Opaque decision-making process: Existing fault diagnosis systems often lack transparency, making it difficult for users to understand and trust their decision-making process, affecting the system's acceptance and usage effectiveness.

[0011] In view of the above defects, the present invention aims to propose a database fault diagnosis and decision-making method, aiming to overcome the limitations of the prior art and achieve a more efficient, accurate and transparent fault diagnosis and decision-making process. Summary of the invention

[0012] The purpose of the present invention is to provide a database fault diagnosis decision method, system and readable storage medium, which at least solves one of the key problems in existing database fault diagnosis technology, such as poor dynamic adaptability, excessive reliance on manual labor, lagging knowledge update and opaque decision-making process.

[0013] In order to achieve the above object, the present invention provides a database fault diagnosis decision method, comprising the following steps:

[0014] Build a decision-making center based on a large language model;

[0015] Utilizing the decision center to receive fault information and perform analysis;

[0016] Based on the parsed information, a dynamic decision tree is constructed through the decision center, and the dynamic decision tree can dynamically adjust its branches and decision points according to known fault modes and real-time fault information;

[0017] The task is distributed to a multi-intelligent agent network through the decision center. The multi-intelligent agent network is configured with multiple intelligent agents, each of which focuses on a specific type of fault diagnosis task. Each intelligent agent performs the assigned task and feeds back the result to the decision center;

[0018] The decision center updates the dynamic decision tree according to the results fed back by each intelligent agent and generates a fault diagnosis report.

[0019] Optionally, after the intelligent agent completes the assigned task, the database fault diagnosis decision method further includes:

[0020] An adversarial verification is performed between the decision center and the intelligent agent, wherein the decision center questions the diagnosis result of the intelligent agent and requires the intelligent agent to provide evidence or further analysis to confirm the accuracy of the result.

[0021] Optionally, the method for constructing a dynamic decision tree through the decision center specifically includes:

[0022] Initializing the dynamic decision tree;

[0023] Use retrieval enhancement generation technology to retrieve past database fault information;

[0024] determining key features for fault diagnosis based on the retrieved information;

[0025] For each key feature, the dynamic decision tree will create a branch to inquire or check the status of the key feature, and adjust the dynamic decision tree according to the result of the inquiry or check.

[0026] Optionally, the method for constructing a dynamic decision tree through the decision center further includes:

[0027] Nodes that require human expert intervention are marked in the dynamic decision tree.

[0028] Optionally, the decision center updates the dynamic decision tree according to the results fed back by each intelligent agent and the opinions of the human experts.

[0029] Optionally, the database fault diagnosis decision method further includes: establishing a transparent feedback mechanism in the decision center, through which a user can track the progress of fault diagnosis and view the status of the dynamic decision tree.

[0030] Optionally, after the dynamic decision tree is generated, the database fault diagnosis decision method further includes:

[0031] A decision tree evaluation model is introduced to evaluate the results.

[0032] Optionally, during the training process of the decision tree evaluation model, the pre-trained language model is fine-tuned using the LoRA method. The specific training process is as follows:

[0033] Let M be the pre-trained language model, L be the loss function, and D be the historical fault case dataset. Under the LoRA framework, only the inserted low-rank matrix is ​​updated. , while keeping the main model parameters θ unchanged, the training objective is expressed as:

[0034]

[0035] Among them, x is the input case and y is the corresponding expert opinion;

[0036] Through iterative optimization, the difference between the decision tree evaluation results and the expert opinions is sought to be minimized.

[0037] Based on the same inventive concept, the present invention also provides a database fault diagnosis decision system, comprising:

[0038] A decision center, built based on a large language model, configured to receive and parse fault information, and to build a dynamic decision tree based on the parsed information, wherein the dynamic decision tree can dynamically adjust its branches and decision points according to known fault modes and real-time fault information;

[0039] A multi-intelligent agent network is configured to execute the tasks assigned by the decision center and feed back the results to the decision center, wherein the multi-intelligent agent network is configured with multiple intelligent agents, each of which focuses on a specific type of fault diagnosis task;

[0040] The decision center is also configured to update the dynamic decision tree according to the results fed back by each intelligent agent and generate a fault diagnosis report.

[0041] Based on the same inventive concept, the present invention also provides a readable storage medium on which a computer program is stored. When the computer program is executed, the database fault diagnosis decision method as described above can be implemented.

[0042] The database fault diagnosis decision method, system and readable storage medium provided by the present invention have at least one of the following beneficial effects:

[0043] 1) By introducing a decision-making center and multi-intelligent agent network based on a large language model, fault diagnosis is highly intelligent and automated, reducing the need for manual intervention and improving the efficiency and accuracy of fault diagnosis;

[0044] 2) Building a dynamic decision tree can quickly adapt to changes in the database system architecture and emerging fault types, enhancing the dynamic adaptability and flexibility of the system;

[0045] 3) By explicitly identifying the nodes that require human experts in the dynamic decision tree, the human-machine collaboration process is optimized, ensuring that the knowledge and experience of human experts can be fully utilized when necessary;

[0046] 4) By establishing a transparent feedback mechanism and visual presentation of the dynamic decision tree, users can be informed of the progress of fault diagnosis at any time and provide additional information when necessary, which promotes effective communication between humans and machines, enhances the transparency of the fault diagnosis process, enables users to understand the diagnostic logic, and improves their trust in the diagnostic results. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Those skilled in the art should understand that the drawings are provided for a better understanding of the present invention and do not constitute any limitation on the scope of the present invention.

[0048] Figure 1 The present invention is a flowchart of a database fault diagnosis and decision-making method provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0049] In order to make the purpose, advantages and features of the present invention clearer, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the accompanying drawings are in a very simplified form and use non-precise proportions, which are only used to conveniently and clearly assist in explaining the purpose of the embodiments of the present invention. In order to make the purpose, features and advantages of the present invention more obvious and easy to understand, please refer to the accompanying drawings. It should be noted that the structure, proportion, size, etc. illustrated in the drawings of this specification are only used to match the content disclosed in the specification for people familiar with this technology to understand and read, and are not used to limit the limiting conditions for the implementation of the present invention. Any modification of the structure, change in the proportional relationship or adjustment of the size, under the same or similar conditions as the effects that can be produced by the present invention and the purposes that can be achieved, should still fall within the scope of the technical content disclosed by the present invention.

[0050] Please refer to Figure 1 This embodiment provides a database fault diagnosis decision method, including the following steps:

[0051] S1. Build a decision center based on a large language model;

[0052] S2, using the decision center to receive fault information and perform analysis;

[0053] S3. Based on the parsed information, a dynamic decision tree is constructed through the decision center, and the dynamic decision tree can dynamically adjust its branches and decision points according to known fault modes and real-time fault information;

[0054] S4, distributing the task to a multi-intelligent agent network through the decision center, wherein the multi-intelligent agent network is configured with multiple intelligent agents, each of which focuses on a specific type of fault diagnosis task, and each intelligent agent performs the assigned task and feeds back the result to the decision center;

[0055] S5. The decision center updates the dynamic decision tree according to the results fed back by each intelligent agent and generates a fault diagnosis report.

[0056] By introducing a decision center and multi-intelligent agent network based on a large language model, fault diagnosis is highly intelligent and automated, reducing the need for manual intervention and improving the efficiency and accuracy of fault diagnosis. In addition, building a dynamic decision tree can quickly adapt to changes in the database system architecture and emerging fault types, enhancing the dynamic adaptability and flexibility of the system.

[0057] First, execute S1 to build a decision center based on a large language model. In this embodiment, the decision center is based on an advanced large language model, responsible for receiving user input, parsing fault descriptions, generating the dynamic decision tree, and guiding the subsequent fault diagnosis process according to the dynamic decision tree. The decision center can also dynamically adjust the decision logic to achieve intelligent and automated fault diagnosis.

[0058] Then, S2 is executed to receive and parse the fault information using the decision center. In this embodiment, a user may submit a fault report or description, such as fault symptoms, environmental information, system log fragments, etc., to the decision center, and the decision center parses the user input based on its advanced large language model to understand the nature and context of the fault.

[0059] Then, S3 is executed to construct a dynamic decision tree through the decision center based on the parsed information. The dynamic decision tree can dynamically adjust its branches and decision points according to known fault modes and real-time fault information.

[0060] Furthermore, the method of constructing a dynamic decision tree through the decision center specifically includes:

[0061] S31, initializing the dynamic decision tree;

[0062] S32, using the retrieval enhancement generation technology to retrieve past database fault information;

[0063] S33, determining key features for fault diagnosis based on the retrieved information;

[0064] S34. For each key feature, the dynamic decision tree will create a branch to inquire or check the status of the key feature, and adjust the dynamic decision tree according to the result of the inquiry or check.

[0065] In this embodiment, S31 is first executed, and the dynamic decision tree generates a meta-decision tree using a dynamic decision tree algorithm, and its generation first starts from a root node, which represents the starting point of fault diagnosis. The root node may ask some general questions, such as "Which system component did the fault occur in?" or "Is there any obvious error information?".

[0066] Then, S32 is executed, and using the retrieval enhancement generation technology (RAG), the decision center will retrieve relevant information such as past database fault tickets, maintenance records, technical documents, forum discussions, etc. This step aims to find cases similar to the current fault description so as to learn how to handle them.

[0067] Then, S3 is executed. Based on the retrieved information, the decision center will determine which features or attributes are most critical for diagnosing the fault (ie, key features), which may include error codes, hardware status, software versions, system log keywords, etc.

[0068] Then execute S4, for each key feature, the dynamic decision tree will create a branch to ask or check the status of this feature. For example, "Is the error code X?" If the answer is yes, the dynamic decision tree will continue to extend downward, otherwise it will turn to another branch.

[0069] Preferably, during the generation of the dynamic decision tree, the decision center will adjust the structure of the dynamic decision tree based on real-time feedback and input from the intelligent agent. This means that if a path proves to be invalid, or new information indicates that there is a better path, the dynamic decision tree will be reconstructed.

[0070] Preferably, during the generation and adjustment of the dynamic decision tree, the search enhancement generation technology is continuously used to ensure that the nodes and branches of the dynamic decision tree are based on the most relevant and latest information. If the search enhancement generation technology finds a more detailed fault ticket or solution at a certain node, the dynamic decision tree may add a new branch or modify an existing path.

[0071] In this embodiment, the method for constructing a dynamic decision tree by the decision center further includes:

[0072] Nodes that require human expert intervention are marked in the dynamic decision tree.

[0073] The decision center will mark the nodes that require human expert intervention, usually when the dynamic decision tree encounters uncertain or complex situations, which may be due to lack of sufficient information. These nodes that require human expert intervention will be fed back to the user in this round of answers. If some node marks in the dynamic decision tree require human expert intervention, the system will prompt the user or relevant experts to make manual judgments, while the intelligent agent continues to automatically process those parts that do not require human intervention, realizing the optimization of human-machine collaboration.

[0074] Preferably, after the dynamic decision tree is generated, the database fault diagnosis decision method further includes:

[0075] A decision tree evaluation model is introduced to evaluate the results.

[0076] Furthermore, during the training process of the decision tree evaluation model, the pre-trained language model is fine-tuned using the LoRA method. The specific training process is as follows:

[0077] Let M be the pre-trained language model, L be the loss function, and D be the historical fault case dataset. Under the LoRA framework, only the inserted low-rank matrix is ​​updated. , while keeping the main model parameters θ unchanged, the training objective is expressed as:

[0078]

[0079] Among them, x is the input case and y is the corresponding expert opinion;

[0080] Through iterative optimization, the difference between the decision tree evaluation results and the expert opinions is sought to be minimized.

[0081] After the dynamic decision tree is constructed, S4 is executed to distribute the tasks to the multi-intelligent agent network through the decision center. The multi-intelligent agent network is configured with multiple intelligent agents, each of which focuses on a specific type of fault diagnosis task, such as data retrieval, performance analysis, log interpretation, etc. The intelligent agent works closely with the decision center and performs the tasks assigned by the decision center, and then feeds the results back to the decision center.

[0082] Preferably, after the intelligent agent completes the assigned task, the database fault diagnosis decision method further includes:

[0083] The decision center conducts adversarial verification with the intelligent agent, where the decision center questions the diagnosis result of the intelligent agent and requires the intelligent agent to provide evidence or further analysis, and confirms that the result returned by the intelligent agent is accurate through a continuous questioning and verification process. In this process, the intelligent agent may need to perform multiple iterations until sufficient information is obtained.

[0084] Finally, S5 is executed, and the decision center updates the dynamic decision tree according to the feedback results of each intelligent agent, adjusts the diagnosis path, optimizes the fault diagnosis process, and generates a fault diagnosis report.

[0085] Preferably, when there are nodes in the dynamic decision tree that require human expert intervention, the decision center will update the dynamic decision tree based on the feedback results of each intelligent agent and the opinions of the human experts. The decision center combines the feedback of all intelligent agents and the opinions of human experts to generate a final fault diagnosis report, including the cause of the problem, recommended solutions and preventive measures, etc. Based on the diagnosis report, the user can execute the recommended solution to fix the fault. After the fault diagnosis is completed, the system records the entire process and adds it to the knowledge base to optimize the decision-making ability of the decision center.

[0086] Preferably, the database fault diagnosis decision method further includes:

[0087] A transparent feedback mechanism is established in the decision center, through which users can track the progress of fault diagnosis and view the status of the dynamic decision tree, so as to understand the process of fault diagnosis and provide additional information when necessary. By establishing a transparent feedback mechanism and visualizing the dynamic decision tree, users can be informed of the progress of fault diagnosis at any time and provide additional information when necessary, thereby promoting effective communication between humans and machines, enhancing the transparency of the fault diagnosis process, enabling users to understand the diagnostic logic, and improving trust in the diagnostic results.

[0088] Based on the same inventive concept, an embodiment of the present invention further provides a database fault diagnosis decision system, including:

[0089] A decision center, built based on a large language model, configured to receive and parse fault information, and to build a dynamic decision tree based on the parsed information, wherein the dynamic decision tree can dynamically adjust its branches and decision points according to known fault modes and real-time data;

[0090] A multi-intelligent agent network is configured to execute the tasks assigned by the decision center and feed back the results to the decision center, wherein the multi-intelligent agent network is configured with multiple intelligent agents, each of which focuses on a specific type of fault diagnosis task;

[0091] The decision center is also configured to update the dynamic decision tree according to the results fed back by each intelligent agent and generate a fault diagnosis report.

[0092] Based on the same inventive concept, an embodiment of the present invention further provides a readable storage medium on which a computer program is stored. When the computer program is executed, the database fault diagnosis decision method as described above can be implemented.

[0093] The readable storage medium can be a tangible device that can keep and store the instructions used by the instruction execution device, such as but not limited to an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device or any suitable combination thereof. The more specific example (non-exhaustive list) of the readable storage medium includes: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a static random access memory (SRAM), a portable compact disk read-only memory (CD-ROM), a digital versatile disk (DVD), a memory stick, a floppy disk, a mechanical encoding device, such as a punch card or a convex structure in a groove on which instructions are stored, and any suitable combination thereof. The computer program described herein can be downloaded from the readable storage medium to each computing / processing device, or downloaded to an external computer or external storage device via a network, such as the Internet, a local area network, a wide area network and / or a wireless network. The network can include copper transmission cables, optical fiber transmission, wireless transmission, routers, firewalls, switches, gateway computers and / or edge servers. The network adapter card or network interface in each computing / processing device receives the computer program from the network and forwards the computer program for storage in the readable storage medium in each computing / processing device. The computer program for performing the operation of the present invention can be an assembly instruction, an instruction set architecture (ISA) instruction, a machine instruction, a machine-related instruction, a microcode, a firmware instruction, a state setting data, or a source code or object code written in any combination of one or more programming languages, including object-oriented programming languages-such as Smalltalk, C++, etc., and conventional procedural programming languages-such as "C" language or similar programming languages. The computer program can be executed completely on the user's computer, partially on the user's computer, as an independent software package, partially on the user's computer and partially on the remote computer, or completely on the remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computer (for example, using an Internet service provider to connect through the Internet). In some embodiments, by utilizing the state information of a computer program to personalize an electronic circuit, such as a programmable logic circuit, a field programmable gate array (FPGA), or a programmable logic array (PLA), the electronic circuit can execute computer-readable program instructions to implement various aspects of the present invention.

[0094] Here, various aspects of the present invention are described with reference to the flowchart and / or block diagram of the method, system and computer program product according to the embodiment of the present invention. It should be understood that each square frame of the flowchart and / or block diagram and the combination of the square frames in the flowchart and / or block diagram can be realized by a computer program. These computer programs can be provided to the processor of a general-purpose computer, a special-purpose computer or other programmable data processing device, so as to produce a machine, so that these programs, when executed by the processor of a computer or other programmable data processing device, produce a device for realizing the function / action specified in one or more square frames in the flowchart and / or block diagram. These computer programs can also be stored in a readable storage medium, and these computer programs make the computer, programmable data processing device and / or other equipment work in a specific way, so that the readable storage medium storing the computer program includes a manufactured product, which includes instructions for realizing various aspects of the function / action specified in one or more square frames in the flowchart and / or block diagram.

[0095] The computer program may also be loaded onto a computer, other programmable data processing apparatus, or other device so that a series of operational steps are executed on the computer, other programmable data processing apparatus, or other device to produce a computer-implemented process, thereby causing the computer program executed on the computer, other programmable data processing apparatus, or other device to implement the functions / actions specified in one or more boxes in the flowchart and / or block diagram.

[0096] In summary, the embodiments of the present invention provide a database fault diagnosis decision method, system and readable storage medium. By introducing a decision center and a multi-intelligent agent network established based on a large language model, fault diagnosis is highly intelligent and automated, the need for manual intervention is reduced, and the efficiency and accuracy of fault diagnosis are improved; by constructing a dynamic decision tree, it can quickly adapt to changes in the database system architecture and emerging fault types, thereby enhancing the dynamic adaptability and flexibility of the system; by clearly identifying the nodes that require human experts in the dynamic decision tree, the human-computer collaboration process is optimized, ensuring that the knowledge and experience of human experts can be fully utilized when necessary; and by establishing a transparent feedback mechanism and a visual presentation of the dynamic decision tree, the user is allowed to understand the progress of the fault diagnosis at any time and provide additional information when necessary, thereby promoting effective communication between man and machine, enhancing the transparency of the fault diagnosis process, enabling users to understand the diagnosis logic, and improving trust in the diagnosis results.

[0097] In addition, it should be recognized that although the present invention has been disclosed as a preferred embodiment, the above embodiment is not intended to limit the present invention. For any technician familiar with the art, without departing from the scope of the technical solution of the present invention, the technical content disclosed above can be used to make many possible changes and modifications to the technical solution of the present invention, or modified into equivalent embodiments of equivalent changes. Therefore, any simple modification, equivalent change and modification made to the above embodiment according to the technical essence of the present invention without departing from the content of the technical solution of the present invention still belongs to the scope of protection of the technical solution of the present invention.

Claims

1. A database fault diagnosis decision method, characterized in that: The following steps are involved: Build a decision-making center based on a large language model; Utilizing the decision center to receive fault information and perform analysis; Based on the parsed information, a dynamic decision tree is constructed through the decision center, wherein the dynamic decision tree can dynamically adjust its branches and decision points according to known fault modes and real-time fault information, and mark nodes in the dynamic decision tree that require human expert intervention; The task is distributed to a multi-intelligent agent network through the decision center. The multi-intelligent agent network is configured with multiple intelligent agents, each of which focuses on a specific type of fault diagnosis task. Each intelligent agent performs the assigned task and feeds back the result to the decision center; The decision center updates the dynamic decision tree according to the feedback results of each intelligent agent and the opinions of the human experts, and generates a fault diagnosis report; The method of constructing a dynamic decision tree through the decision center specifically includes: Initializing the dynamic decision tree; Use retrieval enhancement generation technology to retrieve past database fault information; determining key features for fault diagnosis based on the retrieved information; For each key feature, the dynamic decision tree will create a branch to inquire or check the status of the key feature, and adjust the dynamic decision tree according to the result of the inquiry or check.

2. The database fault diagnosis decision method according to claim 1, characterized in that: After the intelligent agent completes the assigned task, the database fault diagnosis decision method further includes: An adversarial verification is performed between the decision center and the intelligent agent, wherein the decision center questions the diagnosis result of the intelligent agent and requires the intelligent agent to provide evidence or further analysis to confirm the accuracy of the result.

3. The database fault diagnosis decision method according to claim 1, characterized in that: The database fault diagnosis decision method further includes: establishing a transparent feedback mechanism in the decision center, through which a user can track the progress of fault diagnosis and view the status of the dynamic decision tree.

4. The database fault diagnosis decision method according to claim 1, characterized in that: After the dynamic decision tree is generated, the database fault diagnosis decision method further includes: A decision tree evaluation model is introduced to evaluate the results.

5. The database fault diagnosis decision method according to claim 4, characterized in that: During the training process of the decision tree evaluation model, the LoRA method is used to fine-tune the pre-trained language model. The training process is as follows: Let M be the pre-trained language model, L be the loss function, and D be the historical fault case dataset. Under the LoRA framework, only the inserted low-rank matrix is ​​updated. , while keeping the main model parameters θ unchanged, the training objective is expressed as: Among them, x is the input case and y is the corresponding expert opinion; Through iterative optimization, the difference between the decision tree evaluation results and the expert opinions is sought to be minimized.

6. A database fault diagnosis decision system, characterized in that: include: A decision center, built based on a large language model, configured to receive and parse fault information, and to build a dynamic decision tree based on the parsed information, wherein the dynamic decision tree can dynamically adjust its branches and decision points according to known fault modes and real-time fault information, and mark nodes in the dynamic decision tree that require human expert intervention; A multi-intelligent agent network is configured to execute the tasks assigned by the decision center and feed back the results to the decision center, wherein the multi-intelligent agent network is configured with multiple intelligent agents, each of which focuses on a specific type of fault diagnosis task; The decision center is also configured to update the dynamic decision tree according to the results fed back by each intelligent agent and generate a fault diagnosis report; The method of constructing a dynamic decision tree through the decision center specifically includes: Initializing the dynamic decision tree; Use retrieval enhancement generation technology to retrieve past database fault information; determining key features for fault diagnosis based on the retrieved information; For each key feature, the dynamic decision tree will create a branch to inquire or check the status of the key feature, and adjust the dynamic decision tree according to the result of the inquiry or check.

7. A readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed, it can implement the database fault diagnosis decision method according to any one of claims 1 to 5.

Images