Equipment fault diagnosis method and system based on priori knowledge and data driving

Abstract

The invention discloses an equipment fault diagnosis method and system based on prior knowledge and data driving. The method comprises the following steps: inputting real-time data of equipment operation into an expert rule base for analysis and matching, and determining the category of a fault event and a first confidence coefficient; inputting real-time data of equipment operation into the fault event prediction model to obtain probability distribution and second confidence of various fault events; for the same fault event, based on the first confidence coefficient and the second confidence coefficient, determining the weight of the first confidence coefficient and the fault event probability of the fault event; and weighting the first confidence and the fault event probability of the same fault event to obtain a third confidence of the fault event, and outputting the fault event with the highest third confidence as a diagnosis result. According to the method, expert priori knowledge and the prediction result of the data-driven model can be dynamically and adaptively fused, so that the accuracy and reliability of diagnosis are improved.

Description

Technical Field

[0001] This application relates to the field of industrial equipment fault prediction and health management technology, specifically to a method and system for equipment fault diagnosis based on prior knowledge and data-driven approaches. Background Technology

[0002] In modern industry, equipment fault diagnosis mainly relies on two technical approaches: expert rule-based and data-driven predictive models. Each approach has its advantages and disadvantages, together forming the core foundation of fault diagnosis technology. Expert rule-based methods suffer from drawbacks such as high costs in building and maintaining rule bases, limited fault coverage, and a lack of self-learning capabilities. Data-driven methods, on the other hand, depend on large-scale, high-quality data, and exhibit instability and poor interpretability when data is scarce or operating conditions change abruptly.

[0003] Existing technologies attempt to simply combine two methods: using an expert rule system for initial screening followed by a predictive model for fine-tuning, or using the output of the rule system as input features for the predictive model, to achieve a preliminary fusion of the two technologies. However, existing fusion schemes are static and fixed, failing to dynamically adjust the contribution weights of expert rules and the predictive model in the final decision based on specific fault scenarios, data confidence levels, and model output uncertainty. This results in poor diagnostic performance in complex scenarios such as when rules apply but the model misjudges, or when the model discovers new features that the rules do not cover, failing to meet the needs of adaptive collaborative diagnosis. Summary of the Invention

[0004] The purpose of this application is to address the problem that existing technologies suffer from static and fixed fusion modes, failing to dynamically adjust the decision weights of expert rules and prediction models based on real-time equipment operating status, leading to poor diagnostic performance in complex scenarios. This application provides a device fault diagnosis method and system based on prior knowledge and data-driven approaches. This application can dynamically and adaptively fuse the prediction results of expert prior knowledge and data-driven models, thereby improving the accuracy and reliability of diagnosis.

[0005] To achieve the above objectives, this application provides the following technical solution: The first aspect of this application provides a device fault diagnosis method based on prior knowledge and data-driven approaches, including: Real-time data on equipment operation is input into an expert rule base for analysis and matching to determine the category of fault events and their first confidence level. By inputting real-time data of equipment operation into the fault event prediction model, the probability distribution and second confidence level of various fault events are obtained; For the same fault event, the weights of the first confidence level and the probability of the fault event are determined based on the first confidence level and the second confidence level. The third confidence level of a fault event is obtained by weighting the first confidence level and the probability of the fault event. The fault event with the highest third confidence level is output as the diagnostic result.

[0006] Furthermore, the expert rule base is constructed in the following way: Typical failure cases identified by experts are transformed into structured data. The content of typical failure cases includes time, process, cause and handling measures. For each piece of structured data, a first confidence level is assigned according to a rule; the first confidence level is calculated based on the historical accuracy of the fault event and the expert weight.

[0007] Furthermore, the fault event prediction model is constructed in the following manner: A fault event prediction model is obtained by training a pre-set deep learning model based on the historical data of equipment operation. The input data consists of historical data on equipment operation, including vibration, temperature, and pressure during equipment operation. The output is the probability distribution of various fault events.

[0008] Furthermore, the second confidence level is jointly determined by the classification confidence level output by the fault event prediction model and the accuracy of the fault event prediction model's predictions.

[0009] Furthermore, the method of weighting the first confidence level and the probability of the same fault event to obtain the third confidence level of the fault event is as follows:

[0010] In the formula, Indicates the third confidence level of the failure event; Indicates the first confidence level of the failure event; The weight representing the first confidence level; This represents the probability of a failure event.

[0011] Furthermore, for the same fault event, the weights of the first confidence level and the probability of the fault event are determined based on the first confidence level and the second confidence level of the fault event. The factors for calculating the weights include: the first confidence level output by the expert rule base, the second confidence level of the fault event prediction model, real-time data quality, and fault type characteristics. A second aspect of this application provides a device fault diagnosis system based on prior knowledge and data-driven methods, comprising: The fault rule judgment module is used to input real-time data of equipment operation into the expert rule base for analysis and matching to determine the category of fault events and the first confidence level. The fault model prediction module is used to input real-time data of equipment operation into the fault event prediction model to obtain the probability distribution and second confidence level of various fault events; The weight calculation module is used to determine the weights of the first confidence level and the probability of the fault event for the same fault event, based on the first confidence level and the second confidence level. The diagnostic result output module is used to weight the first confidence level and the probability of the same fault event to obtain the third confidence level of the fault event, and the fault event with the highest third confidence level is output as the diagnostic result.

[0012] A third aspect of this application provides an electronic device, characterized in that it includes: a memory and one or more processors; the memory is coupled to the processors; wherein the memory stores computer program code, the computer program code including computer instructions, and when the computer instructions are executed by the processor, the electronic device performs the steps of the above-described device fault diagnosis method based on prior knowledge and data-driven methods.

[0013] A fourth aspect of this application provides a computer-readable storage medium, characterized in that the computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the above-described device fault diagnosis method based on prior knowledge and data-driven methods.

[0014] The fifth aspect of this application provides a computer program product, including a computer program / instructions, characterized in that, when the computer program / instructions are executed by a processor, they implement the steps of the above-described device fault diagnosis method based on prior knowledge and data-driven methods.

[0015] Compared with the prior art, this application has the following beneficial technical effects: This application calculates a third confidence level and determines the diagnostic result by weighting the first confidence level output from the expert rule base, the failure probability output from the prediction model, and the second confidence level for the same fault event. This enables adaptive fusion of expert prior knowledge and data-driven model prediction results, effectively avoiding the limitations of single expert rules or single prediction models in complex scenarios and significantly improving the accuracy and reliability of equipment fault diagnosis. Simultaneously, the fusion process comprehensively considers factors, including the matching basis of the expert rule base, the probability distribution results of the prediction model, and various confidence level values, meeting the stringent requirements for interpretability of fault diagnosis decisions in industrial settings. This application employs a dynamic weight adjustment mechanism, rather than mechanically executing a fixed fusion process. Instead, it flexibly adjusts the weight allocation strategy based on actual operating conditions, data confidence levels, and model output characteristics. Furthermore, it records the optimal weight parameters under different operating conditions, effectively accumulating diagnostic experience and providing data support for subsequent iterative optimization of the diagnostic model. Attached Figure Description

[0016] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a flowchart of the device fault diagnosis method based on prior knowledge and data-driven approach provided in this application; Figure 2 This is a schematic diagram of the structure of a preferred embodiment of the device fault diagnosis system based on prior knowledge and data-driven methods in this application. Figure 3 This is a schematic diagram of the electronic device structure according to a preferred embodiment of this application. Detailed Implementation

[0018] Embodiments of this disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of this disclosure. It should be understood that the accompanying drawings and embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of this disclosure.

[0019] It should be understood that the steps described in the method embodiments of this disclosure may be performed in different orders and / or in parallel. Furthermore, the method embodiments may include additional steps and / or omit the steps shown. The scope of this disclosure is not limited in this respect.

[0020] The term "comprising" and its variations as used herein are open-ended, meaning "including but not limited to". The term "based on" means "at least partially based on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". Definitions of other terms will be given in the description below. It should be noted that the concepts of "first", "second", etc., used in this disclosure are only used to distinguish different devices, modules, or units, and are not intended to limit the order of functions performed by these devices, modules, or units or their interdependencies.

[0021] It should be noted that the terms "a" and "a plurality of" used in this disclosure are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0022] The names of messages or information exchanged between multiple devices in the embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of such messages or information.

[0023] Traditional equipment fault diagnosis methods, whether based on expert rules or data-driven models, have limitations. Expert rule-based methods are costly to build and maintain, and struggle to cover complex or non-existent faults; while data-driven models heavily rely on high-quality data and suffer from insufficient interpretability of the decision-making process. Existing technologies attempt to combine the two, but their weight adjustments are static and fixed, failing to dynamically adjust based on specific fault scenarios, data confidence levels, and the uncertainty of model outputs. This results in insufficient diagnostic performance under complex conditions.

[0024] like Figure 1 As shown, this application provides a device fault diagnosis method based on prior knowledge and data-driven approaches, including: Real-time data on equipment operation is input into an expert rule base for analysis and matching to determine the category of fault events and their first confidence level. By inputting real-time data of equipment operation into the fault event prediction model, the probability distribution and second confidence level of various fault events are obtained; For the same fault event, the weights of the first confidence level and the probability of the fault event are determined based on the first confidence level and the second confidence level. The third confidence level of a fault event is obtained by weighting the first confidence level and the probability of the fault event. The fault event with the highest third confidence level is output as the diagnostic result.

[0025] The method proposed in this application achieves adaptive diagnosis of equipment faults by dynamically fusing the first confidence score output from the expert rule base with the fault event probability output from the fault event prediction model. This overcomes the limitations of traditional methods, which have fixed weights and cannot be adjusted according to real-time diagnostic status. This method effectively balances the interpretability of expert knowledge with the generalization ability of data-driven models, improving the accuracy and robustness of fault diagnosis and reducing the occurrence of misdiagnosis and missed diagnosis when facing complex and ever-changing equipment operating conditions.

[0026] In some specific implementations, the expert rule base is constructed in the following ways: Typical failure cases identified by experts are transformed into structured data. The content of typical failure cases includes time, process, cause and handling measures. A first confidence level is assigned to each piece of structured data. The first confidence level is calculated based on the historical accuracy of the failure event and the expert weight.

[0027] Specifically, transforming typical failure cases identified by experts into structured data refers to organizing and storing representative failure events, identified, analyzed, and recorded by domain experts based on their experience and knowledge, according to predefined formats and fields. The first confidence level represents the degree of trust given by the expert rule base for a specific failure event category. The first confidence level is calculated based on the historical accuracy of the failure event and expert weights, integrating objective data and subjective experience. A higher first confidence level is assigned when real-time data highly matches the expert rules. For example, an example of an expert rule base, rule... If the vibration amplitude X is greater than the threshold A and the frequency components include the characteristic frequency B, then the fault is "rotor imbalance," with a confidence level of [insert confidence level here]. ;rule If the bearing temperature Y is greater than the threshold C, and the lubricating oil temperature rise rate is greater than the threshold D, then the fault is "bearing wear," with a confidence level of [insert confidence level here]. .

[0028] In some specific implementations, the fault event prediction model is constructed in the following way: A fault event prediction model is obtained by training a preset deep learning model based on the historical data of equipment operation; wherein, the input data is the historical data of equipment operation, which includes the vibration, temperature and pressure of the equipment during operation, and the output result is the probability distribution of various fault events.

[0029] Specifically, historical equipment operation data typically originates from sensor data, control system logs, or manually recorded structured or semi-structured data. This historical data usually undergoes preprocessing steps such as cleaning, noise reduction, missing value imputation, and normalization before training to ensure data quality and the effectiveness of model training. The pre-set deep learning model includes one or both of Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU) networks. The process involves iteratively optimizing the parameters of the pre-set deep learning model using the historical equipment operation data. During this process, the model learns the complex nonlinear mapping relationship between input data and fault events to minimize prediction errors.

[0030] In some specific implementations, the second confidence level is jointly determined by the classification confidence level output by the fault event prediction model and the prediction accuracy of the fault event prediction model. The determination of the second confidence level no longer relies solely on the classification confidence level of the fault event prediction model for current real-time data, but further integrates the prediction accuracy demonstrated by the model on historical data. This combined approach allows the second confidence level to more comprehensively and robustly reflect the reliability of the fault event prediction model.

[0031] In some specific implementations, the third confidence level of a fault event is obtained by weighting the first confidence level and the probability of the fault event, specifically as follows:

[0032] In the formula, Indicates the third confidence level of the failure event; Indicates the first confidence level of the failure event; The weight representing the first confidence level; This represents the probability of a failure event.

[0033] In some specific implementations, for the same fault event, the weights of the first confidence level and the probability of the fault event are determined based on the first confidence level and the second confidence level of the fault event. The factors for calculating the weights include: the first confidence level output by the expert rule base, the second confidence level of the fault event prediction model, real-time data quality, and fault type characteristics.

[0034] Specifically, the first confidence level of the expert rule base output: when the expert rule base returns at least one successfully matched fault diagnosis conclusion, and its confidence level is higher than the first preset threshold, the expert rule weight is increased; when the rule base has no matching results or the confidence levels of all matching conclusions are lower than the threshold, the expert rule weight is decreased. The second confidence level of the fault event prediction model: when the second confidence level is lower than the second preset threshold, the expert rule weight is increased; when the second confidence level is higher than the threshold and the fault type corresponding to its highest probability is consistent with the fault type output by the expert rule base, the expert rule weight is decreased. Real-time data quality: when real-time data contains noise, is missing, or significantly exceeds the training set range, the rule matching weight is increased. Fault type characteristics: For typical faults with abundant historical data and clear mechanisms, increase or decrease the weight of expert rules; for faults that have not occurred or are complex, decrease the weight of expert rules.

[0035] Example 1 S1: Real-time diagnostics and confidence assessment. Real-time data D from equipment operation shows that vibration amplitude X slightly exceeds A. The fault event prediction model calculates the probability of "rotor imbalance". The second confidence level is low. The first confidence level of the expert rule base Return to expert rule base knowledge .

[0036] S2: Knowledge fusion weights are adaptively adjusted, calculated based on the following conditions. High-confidence rule knowledge is invoked. Therefore, the weight allocation is biased towards the expert rule base. The fault event prediction model predicts the probability of "rotor imbalance". The second confidence level is low. Both conclusions are consistent. Considering the slight noise in the vibration data, after comprehensive judgment, the following setting is made. .

[0037] S3: Weighted fusion and decision output, third confidence level for "rotor imbalance" fault: The output diagnosis result is "Rotor imbalance," with a final confidence level of 84%. An explanation is generated: "Expert rules contribute 64% to this result, while the prediction model contributes 36%."

[0038] like Figure 2 As shown, another objective of this application is to propose a device fault diagnosis system based on prior knowledge and data-driven methods, comprising: The fault rule judgment module is used to input real-time data of equipment operation into the expert rule base for analysis and matching to determine the category of fault events and the first confidence level. The fault model prediction module is used to input real-time data of equipment operation into the fault event prediction model to obtain the probability distribution and second confidence level of various fault events; The weight calculation module is used to determine the weights of the first confidence level and the probability of the fault event for the same fault event, based on the first confidence level and the second confidence level. The diagnostic result output module is used to weight the first confidence level and the probability of the same fault event to obtain the third confidence level of the fault event, and the fault event with the highest third confidence level is output as the diagnostic result.

[0039] Please see Figure 3 As shown, this application also provides an electronic device 100 for implementing a device fault diagnosis method based on prior knowledge and data-driven methods; the electronic device 100 includes a memory 101, at least one processor 102, a computer program 103 stored in the memory 101 and executable on the at least one processor 102, and at least one communication bus 104.

[0040] The memory 101 can be used to store the computer program 103. The processor 102 implements the steps of the device fault diagnosis method based on prior knowledge and data-driven methods by running or executing the computer program stored in the memory 101 and calling the data stored in the memory 101. The memory 101 may mainly include a program storage area and a data storage area. The program storage area may store the operating system, application programs required for at least one function (such as sound playback function, image playback function, etc.), etc.; the data storage area may store data created according to the use of the electronic device 100 (such as audio data), etc. In addition, the memory 101 may include non-volatile memory, such as hard disk, memory, plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, at least one disk storage device, flash memory device, or other non-volatile solid-state storage device.

[0041] The at least one processor 102 may be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The processor 102 may be a microprocessor or any conventional processor. The processor 102 is the control center of the electronic device 100, connecting various parts of the electronic device 100 via various interfaces and lines.

[0042] The memory 101 in the electronic device 100 stores multiple instructions to implement a device fault diagnosis method based on prior knowledge and data-driven approach, and the processor 102 can execute the multiple instructions to achieve the following: Real-time data on equipment operation is input into an expert rule base for analysis and matching to determine the category of fault events and their first confidence level. By inputting real-time data of equipment operation into the fault event prediction model, the probability distribution and second confidence level of various fault events are obtained; For the same fault event, the weights of the first confidence level and the probability of the fault event are determined based on the first confidence level and the second confidence level. The third confidence level of a fault event is obtained by weighting the first confidence level and the probability of the fault event. The fault event with the highest third confidence level is output as the diagnostic result.

[0043] If the modules / units integrated in the electronic device 100 are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, and a read-only memory (ROM).

[0044] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0045] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0046] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0047] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0048] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application and not to limit them. Although this application has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of this application. Any modifications or equivalent substitutions that do not depart from the spirit and scope of this application should be covered within the protection scope of the claims of this application.

Claims

1. A device fault diagnosis method based on prior knowledge and data-driven approach, characterized in that, include: Real-time data on equipment operation is input into an expert rule base for analysis and matching to determine the category of fault events and their first confidence level. By inputting real-time data of equipment operation into the fault event prediction model, the probability distribution and second confidence level of various fault events are obtained. For the same fault event, the weights of the first confidence level and the probability of the fault event are determined based on the first confidence level and the second confidence level. The third confidence level of a fault event is obtained by weighting the first confidence level and the probability of the fault event. The fault event with the highest third confidence level is output as the diagnostic result.

2. The equipment fault diagnosis method based on prior knowledge and data-driven approach according to claim 1, characterized in that, The expert rule base is constructed in the following way: Typical failure cases identified by experts are transformed into structured data. The content of typical failure cases includes time, process, cause and handling measures. For each piece of structured data, a first confidence level is assigned according to a rule; the first confidence level is calculated based on the historical accuracy of the fault event and the expert weight.

3. The equipment fault diagnosis method based on prior knowledge and data-driven approach according to claim 1, characterized in that, The fault event prediction model is constructed in the following way: A fault event prediction model is obtained by training a pre-set deep learning model based on the historical data of equipment operation. The input data consists of historical data on equipment operation, including vibration, temperature, and pressure during equipment operation. The output is the probability distribution of various fault events.

4. The equipment fault diagnosis method based on prior knowledge and data-driven approach according to claim 1, characterized in that, The second confidence level is jointly determined by the classification confidence level output by the fault event prediction model and the accuracy of the fault event prediction model.

5. The equipment fault diagnosis method based on prior knowledge and data-driven approach according to claim 1, characterized in that, The method of weighting the first confidence level and the probability of the same fault event to obtain the third confidence level of the fault event is as follows: In the formula, Indicates the third confidence level of the failure event; Indicates the first confidence level of the failure event; The weight representing the first confidence level; This represents the probability of a failure event.

6. The equipment fault diagnosis method based on prior knowledge and data-driven approach according to claim 1, characterized in that, For the same fault event, the weights of the first confidence level and the probability of the fault event are determined based on the first confidence level and the second confidence level of the fault event. The factors for calculating the weights include: the first confidence level output by the expert rule base, the second confidence level of the fault event prediction model, real-time data quality, and fault type characteristics.

7. A device fault diagnosis system based on prior knowledge and data-driven approach, characterized in that, include: The fault rule judgment module is used to input real-time data of equipment operation into the expert rule base for analysis and matching to determine the category of fault events and the first confidence level. The fault model prediction module is used to input real-time data of equipment operation into the fault event prediction model to obtain the probability distribution and second confidence level of various fault events; The weight calculation module is used to determine the weights of the first confidence level and the probability of the fault event for the same fault event, based on the first confidence level and the second confidence level. The diagnostic result output module is used to weight the first confidence level and the probability of the same fault event to obtain the third confidence level of the fault event, and the fault event with the highest third confidence level is output as the diagnostic result.

8. An electronic device, characterized in that, include: A memory, one or more processors; the memory is coupled to the processors; wherein the memory stores computer program code, the computer program code including computer instructions, and when the computer instructions are executed by the processor, the electronic device performs the steps of the device fault diagnosis method based on prior knowledge and data-driven methods as described in any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the steps of the device fault diagnosis method based on prior knowledge and data-driven methods as described in any one of claims 1-6.

10. A computer program product comprising a computer program / instructions, characterized in that, When the computer program / instructions are executed by the processor, they implement the steps of the device fault diagnosis method based on prior knowledge and data-driven methods as described in any one of claims 1-6.

Images