Electrical equipment fault diagnosis method and system based on model correction

Abstract

The invention discloses an electrical equipment fault diagnosis method and system based on model correction, and aims to improve the accuracy and reliability of fault diagnosis. The method comprises the following specific steps: firstly, obtaining electrical equipment operation data and environment data; performing feature extraction on the data to generate multi-source feature data, and performing fusion processing to obtain fused data; inputting the fusion data into a machine learning fault diagnosis model trained by historical operation data and environment data, and outputting a fault category and confidence thereof; and if the confidence coefficient is lower than a preset threshold value, correcting the model hyper-parameter, and re-executing the diagnosis process. According to the method, through multi-source data fusion and a model dynamic correction mechanism, accurate identification and confidence coefficient optimization of electrical equipment faults are realized, and the method is suitable for equipment state monitoring and fault early warning in a complex operation environment.

Description

Technical Field

[0001] The present invention belongs to the technical field of electrical equipment fault diagnosis and health management, and relates to an electrical equipment fault diagnosis method and system based on model correction. Background Art

[0002] In modern industrial production, electrical equipment serves as core infrastructure, and its stable operation is crucial to production efficiency and energy supply, directly impacting public safety and sustainable socioeconomic development. However, over the long term, internal components of electrical equipment gradually age due to electrical and thermal stresses. Furthermore, multiple factors, such as drastic changes in humidity and temperature in the external environment and frequent fluctuations in grid load, contribute to the complex and diverse nature of fault types, encompassing insulation aging, mechanical failures, poor electrical contact, and other forms. This undoubtedly complicates fault diagnosis. Currently, traditional fault diagnosis methods, such as experience-based manual diagnosis, rely primarily on the technician's personal experience and expertise, performing troubleshooting through listening, observing, and measuring. This approach is not only time-consuming and labor-intensive, with extremely low diagnostic efficiency, but also, because diagnostic results are significantly influenced by human factors and are highly subjective, it is prone to misdiagnosis or missed diagnosis. Single-parameter threshold judgment methods, on the other hand, simply compare a single operating parameter of the device with a preset threshold to determine the fault. However, when faced with complex faults involving multiple coupled parameters and interrelated features, accurate identification is often difficult, failing to meet actual diagnostic needs. With the rapid development of the Internet of Things and big data technologies, the massive amounts of data generated during electrical equipment operation have opened new opportunities for fault diagnosis. Some research has begun to explore the use of machine learning models to classify and diagnose electrical equipment faults, hoping to leverage the algorithms' powerful learning capabilities to achieve accurate identification. However, in practical applications, these machine learning models commonly suffer from a key issue: their parameters are fixed after training, lacking a mechanism for dynamic adjustment based on the actual diagnostic scenario. This results in an inability to optimize the model's structure and parameter configuration in response to new equipment operating conditions, shifts in data distribution, or complex and changing fault signatures. This makes it difficult to effectively extract and analyze fault signatures, resulting in a significant decrease in diagnostic accuracy. In modern industrial O&M, where electrical equipment reliability is paramount, this lack of diagnostic accuracy severely hinders the safe and stable operation of equipment and fails to meet the growing demand for high-reliability O&M. Summary of the Invention

[0003] The purpose of the present invention is to solve the technical problem of low fault diagnosis accuracy in the prior art and to provide an electrical equipment fault diagnosis method and system based on model correction.

[0004] In order to achieve the above object, the present invention adopts the following technical solutions: A first aspect of the present invention provides a method for diagnosing electrical equipment faults based on model correction, comprising the following steps: S1, obtain electrical equipment operation data and environmental data; S2, extracting features from electrical equipment operation data and environmental data to generate multi-source feature data; fusing the multi-source feature data to generate fused data; S3, inputting the fused data into a fault diagnosis model for fault diagnosis, and outputting a fault category and a confidence level of the fault category; the fault classification model is obtained by training a machine learning model using historical electrical equipment operation data and historical environmental data; S4, when the confidence level of the fault category is less than a preset threshold, modifying the hyperparameters of the fault diagnosis model; S5, return to S3.

[0005] Furthermore, the feature extraction of the electrical equipment operation data and environmental data adopts a wavelet transform method.

[0006] Furthermore, the machine learning model adopts an SVM model.

[0007] Furthermore, the hyperparameters of the fault diagnosis model are modified as follows: Generate an initial population of hyperparameters for the fault diagnosis model; Based on the value of each particle in the initial population, the fault diagnosis model is trained. After the training is completed, the corresponding fitness function value is calculated. The corresponding historical electrical equipment operation data and the corresponding historical environmental data are used as the training data set. Find the particle value corresponding to the optimal fitness function value and modify the hyperparameters of the fault diagnosis model.

[0008] Furthermore, the hyperparameters include penalty factors and kernel parameters.

[0009] Furthermore, the optimization objective function of the particle swarm algorithm is:

[0010] in, is the accuracy of the fault diagnosis model; is the complexity of the fault diagnosis model, and are all weight coefficients.

[0011] A second aspect of the present invention provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the above-mentioned electrical equipment fault diagnosis method based on model correction when executing the computer program.

[0012] A third aspect of the present invention provides a computer-readable storage medium storing a computer program, wherein the computer program, when executed by a processor, implements the above-mentioned electrical equipment fault diagnosis method based on model correction.

[0013] A fourth aspect of the present invention provides a computer program product, which includes computer instructions, and the computer instructions instruct a computer to execute the above-mentioned electrical equipment fault diagnosis method based on model correction.

[0014] A fifth aspect of the present invention provides an electrical equipment fault diagnosis system based on model correction, comprising: Data acquisition module, which acquires electrical equipment operation data and environmental data; The feature extraction module extracts features from electrical equipment operation data and environmental data to generate multi-source feature data; it fuses the multi-source feature data to generate fused data; A fault diagnosis module inputs the fused data into a fault diagnosis model for fault classification, and outputs a fault category and a confidence level for the fault category; the fault diagnosis model is obtained by training a machine learning model using historical electrical equipment operating data and historical environmental data; The model correction module corrects the hyperparameters of the fault diagnosis model when the confidence of the fault category is less than a preset threshold.

[0015] Compared with the prior art, the present invention has the following beneficial effects: This invention discloses a method for diagnosing electrical equipment faults based on model correction. By synchronously collecting operational data and environmental data (such as temperature, humidity, and load), and performing feature extraction and fusion processing, this method comprehensively covers the multi-dimensional factors affecting the equipment's operating status, avoiding diagnostic bias caused by a single data dimension. This method is particularly suitable for scenarios where environmental changes are sensitive to equipment status. When the confidence level of the fault diagnosis result is insufficient, a hyperparameter correction mechanism is automatically triggered. By iteratively optimizing model parameters, the model's ability to identify complex fault patterns is continuously improved, effectively addressing the problem of decreased diagnostic accuracy of traditional static models in scenarios such as equipment aging and fluctuating operating conditions. The fault diagnosis model output includes the fault category and corresponding confidence level, providing a quantitative evaluation standard for the diagnostic results. This allows maintenance personnel to formulate differentiated maintenance strategies based on confidence levels (e.g., direct warnings for high-confidence results and model correction for low-confidence results), thereby improving the efficiency and scientific nature of fault handling. A machine learning model trained on historical data, combined with a dynamic correction mechanism, can adapt to changes in different equipment types, operating conditions, and environmental conditions, reducing manual parameter adjustment costs. This method is particularly suitable for unified fault monitoring and diagnosis of multiple types of electrical equipment (such as transformers and motors). Through continuous iterative optimization of the diagnostic model, we can timely capture the early potential fault characteristics of equipment, combine confidence assessment to achieve early warning of faults, provide data support for preventive maintenance of equipment, reduce the risk of unplanned downtime, and improve the safety and economy of power system operation.

[0016] Furthermore, the present invention adopts a particle swarm algorithm to optimize model hyperparameters (such as the penalty factor and kernel parameters of SVM), taking a weighted combination of diagnostic accuracy (Accuracy) and model complexity (Complexity) as the objective function, taking into account both precision and computational efficiency; iterative training based on historical data sets quickly locates the optimal hyperparameter combination, allowing the model to dynamically adapt to non-steady-state conditions such as equipment aging and load fluctuations, thereby improving the accuracy of fault diagnosis. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments. It should be understood that the following drawings only illustrate certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other relevant drawings can be obtained based on these drawings without paying any creative work.

[0018] Figure 1 This is a flow chart of the electrical equipment fault diagnosis method based on model correction of the present invention. DETAILED DESCRIPTION

[0019] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions of the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings herein can be arranged and designed in various different configurations.

[0020] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention as claimed, but rather merely represents selected embodiments of the present invention. All other embodiments derived by persons of ordinary skill in the art based on the embodiments of the present invention without creative effort shall fall within the scope of protection of the present invention.

[0021] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, it does not need to be further defined or explained in subsequent drawings.

[0022] The present invention is described in further detail below with reference to the accompanying drawings: See also Figure 1 The present invention discloses a method for diagnosing electrical equipment faults based on model correction, comprising the following steps: S1, obtain electrical equipment operation data and environmental data; S2, extracting features from electrical equipment operation data and environmental data to generate multi-source feature data; fusing the multi-source feature data to generate fused data; S3, inputting the fused data into a fault diagnosis model for fault diagnosis, and outputting a fault category and a confidence level of the fault category; the fault classification model is obtained by training a machine learning model using historical electrical equipment operation data and historical environmental data; S4: When the confidence level of the fault category is less than a preset threshold, the hyperparameters of the fault diagnosis model are modified.

[0023] An embodiment of the present invention provides a method for diagnosing electrical equipment faults based on model correction, comprising the following steps: S1 acquires electrical equipment operating and environmental data. For example, a three-axis vibration accelerometer (sampling frequency 12.8kHz) is placed on the surface of the transformer oil tank. The current waveform of the low-voltage current transformer (CT) is simultaneously collected (sampling frequency 5kHz), along with ambient temperature and humidity data. A sliding window method (window length 1s, 50% overlap) is used to denoise the raw signal, removing spikes caused by electromagnetic interference.

[0024] S2, extract features from the electrical equipment operation data and environmental data to generate multi-source feature data; fuse the multi-source feature data to generate fused data; extract features by performing wavelet transform on the electrical equipment operation data and environmental data obtained in S1; in this embodiment, the multi-source data feature fusion adopts a cascade fusion strategy.

[0025] S3, inputting the fused data into a fault diagnosis model to perform fault diagnosis, and outputting the fault category and the confidence level of the fault category; S301, build a fault diagnosis model: The fault diagnosis model in this embodiment adopts the SVM model. The radial basis function (RBF) kernel SVM is initialized, the penalty factor C is set to 1.0, and the kernel parameter γ is set to 0.1. A historical fault dataset (containing 200 samples each of normal state, turn-to-turn short circuit, and core overheating) is used for 5-fold cross-validation training.

[0026] S302: Input the fusion features collected in real time and output the fault category and confidence level. For example, the output result may be: "Core overheating, confidence level 78%."

[0027] S4, when the confidence level of the fault category is less than a preset threshold, modifying the hyperparameters of the fault diagnosis model; When the confidence level is less than 85% (preset threshold), the model is modified by the particle swarm algorithm. First, the population size, maximum number of iterations and inertia weight of the particle swarm algorithm parameters are set; the optimization objective function is:

[0028] in, is the accuracy of the fault diagnosis model; is the complexity of the fault diagnosis model, and The fault diagnosis model is trained based on the value of each particle in the initial population. After training, the corresponding fitness function value is calculated. The training data set is the corresponding historical electrical equipment operation data and the corresponding historical environmental data. At this time, the historical operation data and historical environmental data involved in the training can be selected as small sample data for rapid training.

[0029] Find the particle value corresponding to the optimal fitness function value and modify the hyperparameters of the fault diagnosis model.

[0030] S5, based on the revised fault diagnosis model, re-output the fault category and the confidence level of the fault category; repeat steps S3-S5 until the confidence level meets the requirement.

[0031] The model-correction-based electrical equipment fault diagnosis method of this invention uses wavelet transforms to extract features from electrical equipment operating and environmental data. Leveraging the advantages of wavelet transforms in time-frequency analysis, it effectively captures transient and subtle fault characteristics in the data, transforming complex raw data into more representative multi-source feature data. Compared to traditional methods, this significantly improves the accuracy and completeness of feature extraction, laying a solid data foundation for subsequent fault diagnosis. The SVM model is selected as the underlying machine learning model. SVM excels in small-sample, nonlinear data classification tasks, making it ideal for the limited amount of electrical equipment fault diagnosis data and the nonlinear distribution of fault characteristics. It can quickly and accurately classify the fused data, outputting reliable fault categories and confidence levels, and effectively addressing the complex fault modes of electrical equipment. A particle swarm algorithm is used to generate an initial population of hyperparameters. The fault diagnosis model is trained and optimized based on a fitness function. Hyperparameters (such as the penalty factor and kernel parameter) are modified through continuous iteration to find the optimal particle values. This mechanism enables the model to dynamically adjust parameters based on actual data, significantly improving its adaptability to complex scenarios such as varying operating conditions and equipment aging, addressing the problem of decreased diagnostic accuracy over time in traditional fixed-parameter models. by The particle swarm algorithm optimizes the objective function, taking into account both the accuracy and complexity of the fault diagnosis model. While improving diagnostic accuracy, it effectively controls the risk of model overfitting, avoids the waste of computing resources and decreased generalization capability caused by overly complex models, and achieves the optimal balance between diagnostic performance and resource utilization, ensuring the efficiency and stability of the model in practical applications. When the confidence level of the fault diagnosis result falls below the preset threshold, the model hyperparameter correction is automatically triggered and the diagnostic process is re-executed. This closed-loop mechanism ensures that the model can detect diagnostic errors in a timely manner, continuously improves diagnostic accuracy through parameter optimization, effectively reduces missed diagnoses and misdiagnoses, and provides reliable fault warnings and decision-making support for the safe and stable operation of electrical equipment, significantly reducing equipment failure risks and operation and maintenance costs.

[0032] An embodiment of the present invention provides an electrical equipment fault diagnosis system based on model correction, characterized by comprising: Data acquisition module, which acquires electrical equipment operation data and environmental data; The feature extraction module extracts features from electrical equipment operation data and environmental data to generate multi-source feature data; it fuses the multi-source feature data to generate fused data; A fault diagnosis module inputs the fused data into a fault diagnosis model for fault classification, and outputs a fault category and a confidence level for the fault category; the fault diagnosis model is obtained by training a machine learning model using historical electrical equipment operating data and historical environmental data; The model correction module corrects the hyperparameters of the fault diagnosis model when the confidence of the fault category is less than a preset threshold.

[0033] In another embodiment of the present invention, an electronic device is provided, the electronic device including a processor and a memory, the memory being used to store a computer program, the computer program including program instructions, and the processor being used to execute the program instructions stored in the computer storage medium. The processor may be a central processing unit (CPU), or may be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc., which is the computing core and control core of the terminal, and is suitable for implementing one or more instructions, specifically suitable for loading and executing one or more instructions to implement the corresponding method flow or corresponding function; the processor described in the embodiment of the present invention can be used for the operation of the electrical equipment fault diagnosis method based on model correction.

[0034] In another embodiment of the present invention, a storage medium is provided, specifically a computer-readable storage medium (Memory). The computer-readable storage medium is a memory device in a terminal device, used to store programs and data. It is understood that the computer-readable storage medium herein may include both built-in storage media in the terminal device and, of course, extended storage media supported by the terminal device. It may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. The computer-readable storage medium provides storage space that stores the terminal's operating system. Furthermore, the storage space also stores one or more instructions suitable for being loaded and executed by a processor. These instructions may be one or more computer programs (including program code). It should be noted that more specific examples (a non-exhaustive list) of computer-readable storage media herein include: an electrical connection having one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), optical fiber, a portable compact disk-read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof.

[0035] Computer-readable storage media also include data signals propagated in baseband or as part of a carrier wave, which carry readable program code. Such propagated data signals can take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The readable storage medium can also be any readable medium other than a readable storage medium, which can send, propagate, or transmit programs for use by or in conjunction with an instruction execution system, device, or device. The program code contained on the readable storage medium can be transmitted using any appropriate medium, including but not limited to wireless, wired, optical cable, RF, etc., or any suitable combination of the above.

[0036] The program code for performing the operations of the present invention may be written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Java, C++, and the like, as well as conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user computing device, partially on the user device, as a stand-alone software package, partially on the user computing device and partially on a remote computing device, or entirely on a remote computing device or server. In the case of a remote computing device, the remote computing device may be connected to the user computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0037] The processor may load and execute one or more instructions stored in the computer-readable storage medium to implement the corresponding steps of the electrical equipment fault diagnosis method based on model correction in the above embodiment.

[0038] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Those skilled in the art will readily appreciate that various modifications and variations of the present invention are possible. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

Claims

1. A method for fault diagnosis of electrical equipment based on model correction, characterized in that: The following steps are involved: S1, obtain electrical equipment operation data and environmental data; S2, extracting features from electrical equipment operation data and environmental data to generate multi-source feature data; Fuse multi-source feature data to generate fused data; S3, inputting the fused data into a fault diagnosis model for fault diagnosis, and outputting a fault category and a confidence level of the fault category; the fault classification model is obtained by training a machine learning model using historical electrical equipment operation data and historical environmental data; S4, when the confidence level of the fault category is less than a preset threshold, modifying the hyperparameters of the fault diagnosis model; S5, return to S3.

2. The electrical equipment fault diagnosis method based on model correction according to claim 1, characterized in that: The feature extraction of the electrical equipment operation data and the environmental data adopts the wavelet transform method.

3. The electrical equipment fault diagnosis method based on fault model correction according to claim 1, characterized in that: The machine learning model adopts the SVM model.

4. The electrical equipment fault diagnosis method based on model correction according to claim 1, characterized in that: The hyperparameters of the fault diagnosis model are modified as follows: Generate an initial population of hyperparameters for the fault diagnosis model; Based on the value of each particle in the initial population, the fault diagnosis model is trained. After the training is completed, the corresponding fitness function value is calculated. The corresponding historical electrical equipment operation data and the corresponding historical environmental data are used as the training data set. Find the particle value corresponding to the optimal fitness function value and modify the hyperparameters of the fault diagnosis model.

5. The electrical equipment fault diagnosis method based on model correction according to claim 4, characterized in that: The hyperparameters include penalty factors and kernel parameters.

6. The electrical equipment fault diagnosis method based on model correction according to claim 4, characterized in that: The optimization objective function of the particle swarm algorithm is: in, is the accuracy of the fault diagnosis model; is the complexity of the fault diagnosis model, and are all weight coefficients.

7. An electronic device, characterized in that: The method comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the electrical equipment fault diagnosis method based on model correction according to any one of claims 1 to 6 when executing the computer program.

8. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the electrical equipment fault diagnosis method based on model correction according to any one of claims 1 to 6 is implemented.

9. A computer program product comprising computer instructions, characterized in that: The computer instructions instruct the computer to execute the electrical equipment fault diagnosis method based on model correction according to any one of claims 1 to 6.

10. An electrical equipment fault diagnosis system based on model correction, characterized in that: include: Data acquisition module, which acquires electrical equipment operation data and environmental data; Feature extraction module, which extracts features from electrical equipment operation data and environmental data to generate multi-source feature data; Fuse multi-source feature data to generate fused data; A fault diagnosis module inputs the fused data into a fault diagnosis model for fault classification, and outputs the fault category and the confidence level of the fault category; The fault diagnosis model is obtained by training a machine learning model using historical electrical equipment operation data and historical environmental data; The model correction module corrects the hyperparameters of the fault diagnosis model when the confidence of the fault category is less than a preset threshold.

Images