Vector matching-based industrial equipment fault diagnosis expert online question-answering system and method

Abstract

The invention discloses an industrial equipment fault diagnosis expert online question answering system and method based on vector matching, a user inputs a question description at a client, the system locally or remotely converts the question description into a vector representation and matches the vector representation with a stored question standard description, a question item closest to the question description is determined, and a question item corresponding to the question description is determined; recommending an expert ID which is good at processing the closest problem item; the user sends a remote communication request to an expert corresponding to the recommended expert ID at the client; an expert receives the remote communication request at a client, and remote online communication is established based on websocket and webRTC technologies; the invention aims to provide efficient, accurate and convenient fault diagnosis service through an intelligent expert matching mechanism, a flexible communication mode and perfect record management, so that the overall performance and user experience of fault diagnosis of industrial equipment are improved.

Description

Technical Field

[0001] The present invention belongs to the field of artificial intelligence technology, and in particular relates to an online question-and-answer system and method for industrial equipment fault diagnosis experts based on vector matching. Background Art

[0002] The technologies used in industrial equipment fault diagnosis include traditional expert systems, rule-based diagnostic methods, and machine learning and artificial intelligence technologies. Traditional expert systems mainly rely on the experience and knowledge base of domain experts to perform fault diagnosis through preset rules and logic. The basic principle is to match the fault phenomenon with the possible cause of the fault, and then provide corresponding diagnostic suggestions based on the matching results. The limitation of this method is that it relies on the experience level of experts, it is difficult to handle complex and emerging fault modes, and the diagnostic accuracy and efficiency are limited.

[0003] Rule-based diagnostic methods usually use fixed diagnostic rules and logic to analyze equipment failures. These rules are usually pre-defined based on the operating characteristics and failure modes of the equipment. The system matches the equipment's status parameters to find the cause of the failure that meets the conditions. For example, a series of fault conditions and corresponding treatment measures can be preset. When the equipment status parameters meet a certain fault condition, the system will automatically trigger the corresponding treatment measures. Rule-based diagnostic methods also have certain limitations. The construction and maintenance of its rule base requires a lot of manual intervention, and it is difficult to cope with the dynamic changes in the equipment operating environment and failure modes.

[0004] With the development of artificial intelligence and big data technology, fault diagnosis methods based on machine learning have gradually become a hot topic of research. This method builds a fault diagnosis model by training a large amount of historical fault data to achieve automatic identification and prediction of equipment faults. Machine learning algorithms can handle complex nonlinear relationships, have strong generalization capabilities, and can maintain high diagnostic accuracy when the equipment operating environment and fault mode change. For example, algorithms such as neural networks and support vector machines are widely used in equipment fault diagnosis. By training equipment operating data and building a fault diagnosis model, efficient and accurate fault diagnosis can be achieved, but it is necessary to organize enough reliable fault data or reports.

[0005] In the existing technology, some remote diagnosis systems have been applied to the field of industrial equipment fault management. These systems usually have basic fault reporting, expert matching and online communication functions, but they are limited in intelligent matching mechanism, real-time communication technology, diversified communication methods, record management and feedback mechanism. For example, in terms of communication, the general system supports text and picture communication between users and experts, and some systems may support audio and video calls, but the communication technology of these systems may be based on traditional HTTP polling or SIP protocol, which has the problems of high latency and poor real-time performance.

[0006] These issues can be summarized into the following four aspects:

[0007] 1. Low diagnostic efficiency: Traditional fault diagnosis methods usually require on-site experts to analyze and judge, which is time-consuming and inefficient. Even if remote diagnosis technology is used, it is difficult to find suitable experts to solve problems in time due to the low efficiency of expert matching, which affects the speed of diagnosis.

[0008] 2. Low accuracy: Existing remote diagnosis systems often rely on fixed rules or simple keyword matching when matching experts, which makes it difficult to accurately understand the problems described by users. As a result, the matched experts may not have the most appropriate expertise, affecting the accuracy and effectiveness of the diagnosis.

[0009] 3. Poor flexibility: Existing remote diagnosis systems lack flexibility in communication methods, usually only supporting text or voice communication, and cannot meet the diverse needs of users in different situations. In addition, the system is also relatively rough in managing communication records, and cannot achieve efficient historical query and management.

[0010] 4. Inadequate user experience: In existing technical solutions, users cannot easily view the professional information and historical evaluation of experts during the diagnosis process, and lack reference when selecting experts. At the same time, users' evaluation information of experts is updated lagging behind, which cannot effectively promote experts to improve service quality. Summary of the invention

[0011] In order to overcome the shortcomings of the above-mentioned prior art, the purpose of the present invention is to provide an online question-and-answer system and method for industrial equipment fault diagnosis experts based on vector matching, aiming to provide efficient, accurate and convenient fault diagnosis services through intelligent expert matching mechanism, flexible communication methods and perfect record management, thereby improving the overall performance and user experience of industrial equipment fault diagnosis.

[0012] In order to achieve the above object, the technical solution adopted by the present invention is:

[0013] An online question-and-answer system for industrial equipment fault diagnosis experts based on vector matching, including a system matching expert module and a communication module;

[0014] The system matching expert module includes a problem description library, an expert database, a matching unit and a recommendation unit; the problem description library stores problem items and problem standard descriptions; the expert database stores expert data, and the expert data includes expert IDs and problem items that experts are good at handling; the matching unit is used to convert the input problem description into a vector representation, and match it with the problem standard description to determine the problem item closest to the problem description; the recommendation unit recommends the expert ID who is good at handling the closest problem item from the expert database; the problem item is a failure mode or fault mode of industrial equipment; the problem standard description is a text description of the problem item;

[0015] The communication module is used to realize online communication with the expert corresponding to the recommended expert ID.

[0016] In one embodiment, the input question description is text, image or video, and the matching unit converts the input question description into a vector representation, and constructs a latent Dirichlet allocation (LDA) model to match the vector representation with each question standard description, in the following method:

[0017] First, the text data of historical diagnostic reports are preprocessed;

[0018] Then, the text data of the historical diagnosis reports are trained, and the text of each historical diagnosis report is represented as a probability distribution of a set of topics, where each topic is a probability distribution of a word, corresponding to the question item;

[0019] The vector representation is input into the trained model to obtain its topic distribution, that is, the probability distribution of a group of topics, which reflects the possible related topics, that is, the corresponding question items.

[0020] In one embodiment, the preprocessing includes text data cleaning (removing punctuation marks, etc.), removing stop words (using TF-IDF), and segmenting into words or phrases to construct a document-term matrix (document-term frequency matrix).

[0021] In one embodiment, the text of each historical diagnosis report is represented as a probability distribution of a set of topics, and the implementation method is as follows:

[0022] Using Gibbs sampling, for each word in the document, the probability of generating the word under each topic is calculated, and then the topics are reallocated based on the probability. Based on the new topic allocation results, the distribution of each document and topic is updated, including the document-topic distribution and the topic-word distribution. This is repeated until the change in model parameters is less than the preset threshold or the specified number of iterations is reached.

[0023] In one embodiment, the topic is matched with an expert who is good at handling the corresponding problem item, so that the expert ID who is good at handling the closest problem item is obtained in the matching unit and provided to the recommendation unit.

[0024] In one embodiment, the system also includes a record management module, which saves all communication records, generates a unique session ID for each communication, and introduces automatic labeling and classification technology to automatically organize and label historical communication content through an AI model. The saved communication records are used to train the latent Dirichlet allocation (LDA) model.

[0025] In one embodiment, the expert database introduces expert knowledge base management based on a graph database, and constructs the expert's professional field, historical diagnosis cases, and success rate into a complete graph structure.

[0026] In one embodiment, the communication module is based on websocket and webRTC technologies to achieve instant communication and real-time video calls between users and experts, and adopts a communication optimization method based on edge computing during the communication process to deploy a real-time communication module on edge devices close to users and experts.

[0027] In one embodiment, the system also includes a virtual assistant and a solution database, wherein the solution database stores solutions to failure modes or malfunction modes of industrial equipment. The virtual assistant is a chat robot for industrial scenarios, which pre-processes questions submitted by users and searches the solution database for corresponding solutions based on the closest question items. If so, it is displayed to the user.

[0028] In one embodiment, the system also includes a feedback module based on behavioral analysis, which is used to score and evaluate the communication after the user ends, and to intelligently analyze the behavioral patterns (such as response speed, communication quality, etc.) during the communication between the expert and the user, to provide the user with an objective scoring suggestion reference, and to record the user's evaluation information in a database and update the expert's scoring information.

[0029] The present invention also provides an online question-answering method based on the expert online question-answering system, comprising the following steps:

[0030] The user enters the problem description in the client;

[0031] The system converts the problem description into a vector representation locally or remotely, matches it with the stored problem standard description, determines the problem item closest to the problem description, and recommends the expert ID who is good at handling the closest problem item;

[0032] The user sends a remote communication request to the expert corresponding to the recommended expert ID on the client;

[0033] The expert accepts the remote communication request on the client side and establishes remote online communication.

[0034] Compared with the prior art, the present invention has the following beneficial effects:

[0035] (1) The present invention can convert the problem description submitted by the user into a vector representation and match it with the preset problem standard description. This intelligent matching mechanism can dynamically adjust the matching results according to the actual situation, improve the accuracy and efficiency of fault diagnosis, and reduce the dependence on the experience of a single expert.

[0036] (2) The present invention introduces WebSocket and WebRTC technologies to achieve instant communication and real-time video calls between users and experts. WebSocket provides low-latency full-duplex communication capabilities to ensure the real-time nature of question submission and feedback. WebRTC supports point-to-point media transmission and can perform high-quality audio and video communications even under low-bandwidth conditions, improving communication efficiency and user experience.

[0037] (3) The system of the present invention supports multiple message forms such as text, pictures, and videos. It also supports real-time video calls. Users can choose the most suitable way to describe the problem and communicate with experts, thereby obtaining more accurate diagnosis and guidance, thereby optimizing the user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 It is a schematic diagram of the system structure of the present invention.

[0039] Figure 2 Schematic diagram of a diagnostic system according to an embodiment of the present invention.

[0040] Figure 3 It is a schematic diagram of a diagnostic process of an embodiment of the present invention. DETAILED DESCRIPTION

[0041] The embodiments of the present invention are described in detail below with reference to the accompanying drawings and examples.

[0042] In the field of industrial equipment fault diagnosis, it still mainly relies on on-site analysis and processing by experts, and cannot achieve real-time response and fault processing. There is no good online system to remotely solve these problems in real time. Even if there are some remote systems, they only stay at the level of text and pictures, lacking more intuitive communication methods such as real-time video calls.

[0043] In view of the above-mentioned shortcomings of the prior art, the present invention designs an online question-and-answer system for industrial equipment fault diagnosis experts based on vector matching, and introduces websocket and webRTC technologies to realize remote diagnosis, aiming to solve the shortcomings of the prior art through an intelligent expert matching mechanism, flexible communication methods and perfect record management.

[0044] like Figure 1 As shown, the system of the present invention is mainly composed of two parts: a system matching expert module and a communication module.

[0045] Expert matching is an indispensable part of the system of the present invention. Through the system matching experts, users can find suitable experts according to their needs, which is more flexible and effective.

[0046] In the present invention, the system matching expert module mainly includes a problem description library, an expert database, a matching unit and a recommendation unit. The problem description library stores problem items and the problem standard descriptions corresponding to the problem items, wherein the problem items are failure modes or fault modes of industrial equipment, and the problem standard descriptions are text descriptions of the corresponding problem items. The expert database stores expert data, mainly including expert IDs and problem items that experts are good at handling. Experts and problem items generally correspond to many-to-many, that is, one expert may correspond to multiple problem items, and one problem item may also correspond to multiple experts.

[0047] The matching unit and the recommendation unit are important parts of the expert matching module of the system of the present invention. The matching unit is mainly used to convert the problem description input by the user into a vector representation, and match it with the problem standard description in the problem description library to determine the problem item closest to the user's problem description. The recommendation unit recommends the expert ID who is good at handling the closest problem item from the expert database, and the recommended expert ID can be multiple.

[0048] Communication is a basic component of the system of the present invention. The communicating parties can send messages in the form of text, pictures and videos. In addition, real-time video calls can be made, whereby the two parties can communicate effectively to exchange information to solve problems. The communication module of the present invention is used to realize online communication with the experts corresponding to the recommended expert IDs.

[0049] In terms of hardware, the system of the present invention mainly includes user terminals, expert terminals, servers, etc., and each component is connected through a network. Users and experts access the system through their respective terminals, users submit questions, the system matches experts, and then users and experts communicate in real time, and the system performs record management.

[0050] refer to Figure 2In the expert matching stage, the problem description input by the user can be in various forms such as text, image or video. The matching unit can convert the input problem description into a vector representation according to the keywords in the problem description library. For example, when the input is an image or video, it can be identified through the graphic model to obtain a text vector representation. Then, by calculating the cosine similarity between the text vector representation and the problem standard description vector, the most matching failure mode or fault mode, that is, the problem item, can be determined. The system then retrieves experts who are good at handling the mode from the expert database based on the matched failure mode or fault mode, recommends it to the user and displays it on its terminal interface. This intelligent matching mechanism can dynamically adjust the matching results according to the actual situation, improve the accuracy and efficiency of fault diagnosis, and reduce the dependence on the experience of a single expert.

[0051] Furthermore, in order to improve the matching effect and efficiency, the present invention constructs and trains a latent Dirichlet allocation (LDA) model, and uses the trained model to match the vector representation with the standard description of each problem. Figure 2 , the specific implementation method can be described as follows:

[0052] First, the text data of the historical diagnosis report is preprocessed, including text data cleaning (removing punctuation marks, etc.), removing stop words (using TF-IDF), segmenting into words or phrases, and constructing a document-term matrix (document-term frequency matrix).

[0053] Then, the text data of historical diagnostic reports is trained, and the text of each historical diagnostic report is represented as a probability distribution of a set of topics. Each topic is a probability distribution of a word, corresponding to a problem item, that is, a failure mode or a fault mode. Specifically, Gibbs sampling can be used to calculate the probability of generating the word under each topic for each word in the historical diagnostic report document, and then the topics are reallocated according to the probability. According to the new topic allocation results, the distribution of each document and topic is updated, including the document-topic distribution and the topic-word distribution, and the process is repeated until the change in the model parameters is less than the preset threshold, or the specified number of iterations is reached.

[0054] The vector representation of the user's question description is input into the trained model to obtain its topic distribution, that is, the probability distribution of a group of topics, which reflects the possible related topics, that is, the corresponding question items. The topics can be matched with experts who are good at handling the corresponding question items, so that the expert ID who is good at handling the closest question item is obtained in the matching unit and provided to the recommendation unit.

[0055] Based on this, the present invention can directly recommend appropriate experts according to the topic distribution of the input text, thereby directly omitting the part of calculating the cosine similarity and simplifying the matching process.

[0056] According to the above scheme, the present invention obtains the topic distribution from the problem description submitted by the user by training the latent Dirichlet allocation model, so as to determine the more suitable failure mode. In addition, the expert knowledge base management based on the graph database can be introduced to construct the expert's professional field, historical diagnosis cases, success rate and other information into a complete graph structure, and make recommendations in combination with the expert library based on the graph database. This form of expert database can quickly retrieve experts related to the user's problem description, and has better scalability in a large-scale data environment. Finally, according to the matched failure mode, experts who are good at handling the failure mode are retrieved from the database and recommended to the user. This intelligent matching mechanism can dynamically adjust the matching results according to the actual situation, improve the accuracy and efficiency of fault diagnosis, and reduce the dependence on the experience of a single expert.

[0057] The communication module of the present invention is based on websocket and webRTC technology. WebSocket is a protocol for full-duplex communication on a single TCP connection, which can realize real-time data transmission between the server and the client. When communicating, WebSocket technology is used to realize instant communication between users and systems, users and experts, so that users can submit questions in real time and receive system feedback, ensuring the real-time nature of question submission and feedback. WebSocket provides full-duplex communication capabilities, and the low-latency characteristics ensure the efficiency of the fault diagnosis process. WebRTC is an API that supports web browsers for real-time voice or video conversations. It provides point-to-point media transmission capabilities. In the present invention, it is used to realize real-time video calls and multimedia information transmission between users and experts. Through WebRTC, users and experts can communicate face to face, and experts can view the operating status and fault phenomena of the equipment in real time, providing more accurate diagnosis and guidance. In addition, WebRTC supports point-to-point media transmission, ensuring high-quality audio and video communication under low bandwidth conditions, improving communication efficiency and user experience.

[0058] Furthermore, the present invention can adopt a communication optimization method based on edge computing during the communication process. By deploying real-time communication modules on edge devices close to users and experts, it can reduce delays and network congestion and ensure the real-time and stability of communication. This method is particularly suitable for some industrial sites with poor network conditions.

[0059] Furthermore, the system of the present invention also includes a virtual assistant and a solution database, wherein the solution database stores solutions to failure modes or fault modes of industrial equipment, which are basic and common solutions. The virtual assistant is a chat robot for industrial scenarios, which can pre-process questions submitted by users, and search the solution database for corresponding solutions based on the closest problem items. If so, it will be displayed to the user. Through the virtual assistant, the problems described by the user can be pre-processed, and automated basic diagnostic opinions can be provided to assist experts in their work and improve the user experience.

[0060] refer to Figure 3 The online question-answering method using the system of the present invention mainly includes the following aspects:

[0061] 1. Users and experts enter the online question-and-answer system separately.

[0062] 2. The user enters the problem description in the form of text, picture or video on the client, and the system matches the expert with the problem description. The details are as follows:

[0063] The problem description submitted by the user is converted into a vector representation through the problem description library, and then expert matching is performed using the cosine similarity or LDA model. This intelligent matching mechanism improves the accuracy and efficiency of fault diagnosis and reduces the reliance on the experience of a single expert. This process can be completed locally or remotely.

[0064] 3. The user sends a remote communication request, i.e. a chat request, to the expert corresponding to the recommended expert ID on his client.

[0065] 4. The expert accepts the remote communication request on his client and establishes remote online communication.

[0066] 5. Both parties chat.

[0067] The present invention introduces WebSocket and WebRTC technologies into online communication to achieve instant communication and real-time video calls between users and experts. WebSocket provides low-latency full-duplex communication capabilities to ensure the real-time nature of problem submission and feedback. WebRTC supports point-to-point media transmission and can perform high-quality audio and video communications under low-bandwidth conditions, thereby improving communication efficiency and user experience.

[0068] The present invention supports multiple communication methods such as text, pictures, and videos. Users can choose the most suitable method to describe problems and communicate with experts. The real-time video call function allows experts to intuitively view the status of equipment and fault phenomena, provide more accurate diagnosis and guidance, and optimize the user experience. In addition, it is combined with a virtual assistant (a chatbot for industrial scenarios), which can pre-process problems submitted by users, provide automated basic diagnostic opinions, and reduce the burden on experts.

[0069] 6. The user ends the chat and makes a comment.

[0070] The present invention constructs a systematic record management mechanism that can save all communication records, including text, pictures, videos and other content. A unique session ID is generated for each communication to ensure the integrity and traceability of the communication records. Automatic labeling and classification technology is introduced, and the historical communication content is automatically sorted and labeled through the AI ​​model to facilitate subsequent rapid retrieval and analysis. Users and experts can view historical communication records at any time, which is convenient for subsequent reference and problem solving, and improves the reliability of the system and user satisfaction.

[0071] Based on the system's complete recording mechanism, in addition to the existing historical diagnostic reports in document form, the chat records generated during the remote diagnosis process (including problem descriptions and diagnostic results) can also be used as a data source to optimize the LDA model.

[0072] The present invention applies a feedback system based on behavioral analysis. After the communication is over, the user can score and evaluate the communication. By intelligently analyzing the behavioral patterns (such as response speed, communication quality, etc.) during the communication between the expert and the user, the system provides the user with an objective scoring suggestion reference, rather than relying solely on the user's subjective scoring. Afterwards, the system records the user's evaluation information in the database and updates the expert's scoring information. This mechanism helps the system continuously optimize expert matching and service quality, motivates experts to provide better services, and improves the overall user experience.

[0073] According to the above records, the system of the present invention has good cross-field application potential. It is not only suitable for industrial equipment fault diagnosis, but also can be applied to other fields such as online consultation in medical systems, equipment maintenance in manufacturing, etc. Its versatility and flexibility make the system have broad application prospects and market value.

Claims

1. An online question-answering system for industrial equipment fault diagnosis experts based on vector matching, characterized in that: Including system matching expert module and communication module; The system matching expert module includes a problem description library, an expert database, a matching unit and a recommendation unit; the problem description library stores problem items and problem standard descriptions; the expert database stores expert data, and the expert data includes expert IDs and problem items that experts are good at handling; the matching unit is used to convert the input problem description into a vector representation, and match it with the problem standard description to determine the problem item closest to the problem description; the recommendation unit recommends the expert ID who is good at handling the closest problem item from the expert database; the problem item is a failure mode or fault mode of industrial equipment; the problem standard description is a text description of the problem item; The communication module is used to realize online communication with the expert corresponding to the recommended expert ID.

2. According to claim 1, the industrial equipment fault diagnosis expert online question-answering system based on vector matching is characterized in that: The input problem description is text, image or video. The matching unit converts the input problem description into a vector representation, and constructs a latent Dirichlet allocation model to match the vector representation with each problem standard description, as follows: First, the text data of historical diagnostic reports are preprocessed; Then, the text data of the historical diagnosis reports are trained, and the text of each historical diagnosis report is represented as a probability distribution of a set of topics, where each topic is a probability distribution of a word, corresponding to the question item; The vector representation is input into the trained model to obtain its topic distribution, that is, the probability distribution of a group of topics, which reflects the possible related topics, that is, the corresponding question items.

3. According to claim 2, the industrial equipment fault diagnosis expert online question-answering system based on vector matching is characterized in that: The text of each historical diagnosis report is represented as a probability distribution of a set of topics, and the implementation method is as follows: Using Gibbs sampling, for each word in the document, the probability of generating the word under each topic is calculated, and then the topics are reallocated based on the probability. Based on the new topic allocation results, the distribution of each document and topic is updated, including the document-topic distribution and the topic-word distribution. This process is repeated until the change in model parameters is less than the preset threshold or the specified number of iterations is reached.

4. According to claim 2, the industrial equipment fault diagnosis expert online question-answering system based on vector matching is characterized in that: The topic is matched with an expert who is good at handling the corresponding question item, so that the expert ID who is good at handling the closest question item is obtained in the matching unit and provided to the recommendation unit.

5. According to claim 2, the industrial equipment fault diagnosis expert online question-answering system based on vector matching is characterized in that: The system also includes a record management module, which saves all communication records, generates a unique session ID for each communication, and introduces automatic labeling and classification technology to automatically organize and label historical communication content through an AI model. The saved communication records are used to train the latent Dirichlet allocation model.

6. According to claim 1, the industrial equipment fault diagnosis expert online question-answering system based on vector matching is characterized in that: The expert database introduces expert knowledge base management based on a graph database, and constructs the expert's professional field, historical diagnosis cases, and success rate into a complete graph structure.

7. The online question-answering system for industrial equipment fault diagnosis experts based on vector matching according to claim 1 is characterized in that: The communication module is based on websocket and webRTC technologies to achieve instant communication and real-time video calls between users and experts, and adopts a communication optimization method based on edge computing during the communication process, deploying real-time communication modules on edge devices close to users and experts.

8. The online question-answering system for industrial equipment fault diagnosis experts based on vector matching according to claim 1 is characterized in that: The system also includes a virtual assistant and a solution database. The solution database stores solutions to failure modes or malfunction modes of industrial equipment. The virtual assistant is a chat robot for industrial scenarios. It pre-processes questions submitted by users and searches the solution database for corresponding solutions based on the closest question items. If so, it displays them to the user.

9. The online question-and-answer system for industrial equipment fault diagnosis experts based on vector matching according to claim 1 is characterized in that: The system also includes a feedback module based on behavioral analysis, which is used to score and evaluate the communication after the user ends, and to intelligently analyze the behavioral patterns of the expert and the user during the communication process, to provide the user with objective scoring suggestions and references, and to record the user's evaluation information in the database and update the expert's scoring information.

10. The online question-answering method of the online question-answering system for industrial equipment fault diagnosis experts based on vector matching according to claim 1 is characterized in that: The steps include: The user enters the problem description in the client; The system converts the problem description into a vector representation locally or remotely, matches it with the stored problem standard description, determines the problem item closest to the problem description, and recommends the expert ID who is good at handling the closest problem item; The user sends a remote communication request to the expert corresponding to the recommended expert ID on the client; The expert accepts the remote communication request on the client side and establishes remote online communication.

Images