Electrical primary equipment fault detection method, device, equipment, medium and program

Abstract

The invention provides an electrical primary equipment fault detection method, device, equipment, medium and program, and the method comprises the steps: collecting parameters of different electrical primary equipment, and obtaining a corresponding data set; performing principal component analysis on the data set to obtain dimension-reduced data; constructing a training sample set and a test sample set based on the data after dimension reduction; training the constructed deep learning model by using the training sample set and the test sample set to obtain a prediction model; and performing fault detection on the current electrical primary equipment based on the prediction model. A data set is constructed by collecting parameters of different power devices, data dimension reduction is achieved by means of principal component analysis to eliminate interference information and extract core features, then a sample set is constructed based on dimension-reduced data to train a deep learning model, and finally accurate fault detection of the current power device is achieved through a prediction model obtained through training. The system can adapt to dynamic working conditions of equipment, is high in signal interference resistance, is reliable in data, and can accurately capture early fault signals.

Description

Technical Field

[0001] This invention relates to the field of electrical testing, and more particularly to a method, apparatus, equipment, medium, and procedure for detecting faults in primary electrical equipment. Background Technology

[0002] Primary electrical equipment refers to physical devices (such as generators, transformers, circuit breakers, and power cables) that directly participate in the production, transmission, distribution, and consumption of electrical energy. As a core component of the power system, the stability of its operating state directly affects the safe and reliable power supply of the power grid. Therefore, accurate and real-time fault detection of primary electrical equipment, timely identification of potential faults, and intervention measures are crucial for ensuring the stable operation of the power system, reducing operation and maintenance costs, and avoiding large-scale power outages. With the development of power systems towards intelligence and large-scale deployment, higher requirements are placed on the accuracy, adaptability, and reliability of fault detection technology for primary electrical equipment.

[0003] However, existing electrical primary equipment testing technologies still have many key shortcomings, making it difficult to meet actual operational needs. Specifically: First, testing parameters are mostly fixed preset values, unable to dynamically adapt to the dynamic operating conditions of equipment under different loads, ambient temperatures, and voltage levels. This leads to a mismatch between testing standards and actual operating conditions, resulting in a significant decrease in fault detection accuracy and an inability to accurately identify potential faults caused by changes in operating conditions. Second, the testing process involves multiple types of detection signals, including electromagnetic waves, sound waves, and electrical signals. These signals are prone to cross-interference, reducing the ability to identify valid fault signals and greatly increasing the false fault rate. This not only wastes significant maintenance resources but may also lead to disruptions in normal operations due to misjudgments. Third, the shutdown of operating equipment affects the stability of power supply; fourth, existing detection devices lack an effective real-time calibration mechanism, and are prone to accuracy drift due to environmental corrosion and component aging during long-term operation, resulting in reduced reliability of collected data. Fault judgment based on unreliable data further exacerbates the risk of misjudgment and missed judgment; fifth, early faults in primary electrical equipment are often accompanied by weak nanometer-level partial discharge signals. Existing technology is insufficient to capture such weak signals, and the signals are prone to attenuation during transmission and acquisition, making it difficult to detect early faults in a timely manner and posing a significant risk of missed judgment. Once an early fault develops into a serious fault, it will cause huge economic losses and safety hazards.

[0004] Therefore, there is an urgent need for a method, device, equipment, medium, and procedure for detecting faults in primary electrical equipment to improve the above-mentioned problems. Summary of the Invention

[0005] The purpose of this invention is to provide a method, device, equipment, medium and program for detecting faults in primary electrical equipment, which can adapt to the dynamic operating conditions of the equipment, has strong anti-signal interference ability, reliable data and can accurately capture early fault signals.

[0006] In a first aspect, the present invention provides a method for detecting faults in primary electrical equipment, comprising: Collect parameters from different primary electrical devices to obtain the corresponding datasets; Principal component analysis was performed on the dataset to obtain the dimensionality-reduced data; Training and test sample sets are constructed based on the dimensionality-reduced data; The deep learning model is trained using the training sample set and the test sample set to obtain the prediction model; Fault detection is performed on current primary electrical equipment based on predictive models.

[0007] Optionally, parameters of different primary electrical devices can be collected to obtain the corresponding datasets, including: Collect parameters from different primary electrical equipment, preprocess all parameters, and obtain preprocessed data; Correlation analysis is performed on the preprocessed data to obtain the corresponding dataset; Preprocessing includes handling outliers and handling missing values.

[0008] Optionally, principal component analysis can be performed on the dataset to obtain dimensionality-reduced data, including: Standardize the dataset to obtain standardized data; Calculate the covariance matrix of the standardized data; Eigenvalue decomposition of the covariance matrix yields the magnitudes of the eigenvalues ​​of the principal components; Select the top 5 principal components with the largest eigenvalues ​​based on their magnitude; Construct feature vectors based on the five selected principal components; Projecting the dataset onto the principal components yields the dimensionality-reduced data.

[0009] Optionally, the deep learning model can be trained using the training and testing sample sets to obtain the prediction model, including: The deep learning model is trained using the training and test sample sets to obtain an intermediate prediction model for each target object. The intermediate prediction model for each target object is updated and optimized regularly to obtain the prediction model.

[0010] Optionally, primary electrical equipment includes transformers, circuit breakers, and capacitors; and / or parameters include waveforms, amplitudes, frequencies, electromagnetic waves, sound waves, and optical signals.

[0011] Optionally, the deep learning model is a feedforward neural network; The feedforward neural network consists of two hidden layers. The first hidden layer contains 48 neurons, with batch normalization and L2 regularization, and the second hidden layer contains 24 neurons, with batch normalization and Dropout. The activation function of each neuron in both hidden layers is the ReLU activation function. The output layer of the feedforward neural network is a single neuron layer. The mean squared error is used as the loss function to quantify the difference between the predicted and actual values. The Adam optimizer is used to adjust the weights, and an early stopping strategy is used during training to prevent overfitting.

[0012] Secondly, the present invention provides an apparatus comprising modules / units for performing any of the possible designs described in the first aspect. These modules / units can be implemented in hardware or by hardware executing corresponding software.

[0013] Thirdly, the present invention provides an electronic device including a memory and a processor, wherein the memory stores a program that can run on the processor, and when the program is executed by the processor, the electronic device implements a method for performing any of the possible designs described above.

[0014] Fourthly, the present invention provides a readable storage medium storing a program, which, when executed, implements a method of any possible design of any of the above aspects.

[0015] Fifthly, the present invention provides a computer program product, including a computer program that, when executed by a processor, implements the above-described method.

[0016] The beneficial effects of this invention are as follows: Parameters of different primary electrical equipment are collected to obtain corresponding datasets; principal component analysis is performed on the datasets to obtain dimensionality-reduced data; training and testing sample sets are constructed based on the dimensionality-reduced data; a deep learning model is trained using the training and testing sample sets to obtain a prediction model; and fault detection is performed on the current primary electrical equipment based on the prediction model. By combining principal component analysis with deep learning technology, a dataset is constructed by collecting parameters of different power equipment. Principal component analysis is used to reduce the dimensionality of the data to remove interference information and extract core features. Then, a sample set is constructed based on the dimensionality-reduced data to train a deep learning model. Finally, the trained prediction model is used to achieve accurate fault detection of the current power equipment. This method is adaptable to dynamic operating conditions of equipment, has strong anti-signal interference capabilities, provides reliable data, and can accurately capture early fault signals. Attached Figure Description

[0017] Figure 1 A flowchart illustrating a method provided in an embodiment of the present invention; Figure 2This is a schematic diagram of the structure of a device provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions in the embodiments of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without inventive effort are within the scope of protection of this invention. Unless otherwise defined, the technical or scientific terms used herein should have the ordinary meaning understood by those skilled in the art. The terms "comprising" and similar expressions used herein mean that the element or object preceding the word covers the element or object listed following the word and its equivalents, but do not exclude other elements or objects.

[0019] The technical solutions of the embodiments of the present invention will be described below with reference to the accompanying drawings. In the description of the embodiments of the present invention, the terminology used in the following embodiments is for the purpose of describing specific embodiments only and is not intended to limit the present invention. The singular expressions “a,” “the,” “the,” and “this” are intended to also include expressions such as “one or more,” unless the context clearly indicates otherwise. It should also be understood that in the following embodiments of the present invention, “at least one” and “one or more” refer to one or more (including two). The term “and / or” is used to describe the relationship between related objects, indicating that three relationships can exist; for example, A and / or B can represent: A alone, A and B simultaneously, or B alone, where A and B can be singular or plural. The character “ / ” generally indicates that the preceding and following related objects are in an “or” relationship.

[0020] References to "one embodiment" or "some embodiments" in this specification mean that a particular feature, structure, or characteristic described in connection with that embodiment is included in one or more embodiments of the invention. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," and "in still other embodiments" appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including, but not limited to," unless otherwise specifically emphasized. The term "connection" includes both direct and indirect connections, unless otherwise stated. "First" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated.

[0021] In embodiments of the present invention, "exemplarily" or "for example" are used to indicate that they are examples, illustrations, or descriptions. Any embodiment or design described as "exemplarily" or "for example" in embodiments of the present invention should not be construed as being more preferred or advantageous than other embodiments or design solutions. Rather, the use of "exemplarily" or "for example" is intended to present the relevant concepts in a specific manner.

[0022] like Figure 1 As shown, the present invention provides a method comprising: S101: Collect parameters of different primary electrical equipment to obtain the corresponding dataset.

[0023] In some embodiments, primary electrical equipment includes transformers, circuit breakers, and capacitors.

[0024] In other embodiments, the parameters include waveform, amplitude, frequency, electromagnetic waves, sound waves, and optical signals.

[0025] In some other embodiments, collecting parameters from different primary electrical devices to obtain a corresponding dataset includes: collecting parameters from different primary electrical devices, preprocessing all parameters to obtain preprocessed data; performing correlation analysis on the preprocessed data to obtain a corresponding dataset; the preprocessing includes outlier handling and missing value handling.

[0026] S102, perform principal component analysis on the dataset to obtain the dimensionality-reduced data.

[0027] In some embodiments, performing principal component analysis on the dataset to obtain dimensionality-reduced data includes: standardizing the dataset to obtain standardized data; calculating the covariance matrix of the standardized data; performing eigenvalue decomposition on the covariance matrix to obtain the eigenvalues ​​of the principal components; selecting the top 5 principal components with the largest eigenvalues ​​based on their magnitudes; constructing eigenvectors based on the selected 5 principal components; and projecting the dataset onto the principal components to obtain dimensionality-reduced data.

[0028] S103, construct training and testing sample sets based on the dimensionality-reduced data.

[0029] S104, use the training sample set and the test sample set to train the deep learning model to obtain the prediction model.

[0030] In some embodiments, the deep learning model is a feedforward neural network; the feedforward neural network consists of two hidden layers, and the first hidden layer contains 48 neurons, with batch normalization and L2 regularization, and the second hidden layer contains 24 neurons, with batch normalization and Dropout, and the activation function of each neuron in the two hidden layers is the ReLU activation function.

[0031] In other embodiments, the output layer of the feedforward neural network is a single neuron layer, the mean squared error is used as the loss function to quantify the difference between the predicted and actual values, the Adam optimizer is used to adjust the weights, and an early stopping strategy is used during training to prevent overfitting.

[0032] In some other embodiments, training a deep learning model using a training sample set and a test sample set to obtain a prediction model includes: training the deep learning model using the training sample set and the test sample set to obtain an intermediate prediction model for each target object; and periodically updating and optimizing the intermediate prediction model for each target object to obtain a prediction model.

[0033] S105, based on the prediction model, performs fault detection on the current primary electrical equipment.

[0034] The advantages of this invention are that it utilizes principal component analysis combined with deep learning technology to construct a dataset by collecting parameters from different power equipment. Principal component analysis is used to reduce the dimensionality of the data to remove interference information and extract core features. Then, a sample set is built based on the dimensionality-reduced data to train a deep learning model. Finally, the trained prediction model is used to achieve accurate fault detection of the current power equipment. It can adapt to the dynamic operating conditions of the equipment, has strong anti-signal interference capabilities, reliable data, and can accurately capture early fault signals.

[0035] To facilitate understanding, this embodiment further elaborates on the specific implementation process of the above method in conjunction with a specific application scenario, which includes the following steps: Step 1: Parameter Acquisition, Preprocessing, and Dataset Construction Multi-dimensional parameters (such as voltage / current waveforms, signal amplitude, operating frequency, electromagnetic waves generated by equipment operation, mechanical vibration sound waves, partial discharge light signals, etc.) of various primary electrical equipment (such as transformers, circuit breakers, capacitors, disconnectors, etc.) under different operating conditions (rated load, impact load, ambient temperature fluctuations, etc.) are collected. High-precision sensors and data acquisition terminals are used to achieve real-time and synchronous acquisition, ensuring the integrity and timeliness of parameter acquisition. All collected parameters are preprocessed to obtain preprocessed data. Preprocessing specifically includes outlier handling and missing value handling: Outlier handling uses a combination of the 3σ criterion and box plot method to accurately identify and remove abnormal data caused by sensor failure and electromagnetic interference, avoiding interference from outliers to subsequent analysis; Missing value handling uses a complementary K-nearest neighbor imputation method and linear interpolation method. Linear interpolation is used for short-term missing data, and K-nearest neighbor imputation is used for long-term missing data, preserving the original distribution characteristics of the data to the greatest extent. Correlation analysis was performed on the preprocessed data, using both Pearson and Spearman correlation coefficients to quantify the linear and nonlinear correlations between parameters. Redundant parameters were removed, and core correlation parameters were retained, ultimately constructing a high-quality, low-redundancy dataset.

[0036] Step 2: Data Standardization and Principal Component Dimensionality Reduction The dataset constructed in step 1 is standardized using the Z-score standardization method (i.e., subtracting the mean from the data and then dividing by the standard deviation) to eliminate differences in the units of measurement (such as the unit difference between waveform amplitude and frequency) and the differences in numerical range between different parameters, thus avoiding the influence of units on the model training effect and obtaining standardized data. The covariance matrix of the standardized data is calculated, which can intuitively reflect the degree of linear correlation between the parameter variables. Eigenvalue decomposition is performed on the covariance matrix to obtain the eigenvalues ​​corresponding to each principal component. The larger the eigenvalue, the more original data information is contained in the principal component. The top 5 principal components with the largest eigenvalues ​​are selected based on the eigenvalues ​​(ensuring that the cumulative variance contribution rate is not less than 85% to ensure that core information is not lost). A feature vector matrix is ​​constructed based on the selected 5 principal components to clarify the direction of core feature extraction. The dataset is projected onto the constructed principal component feature vectors to realize the mapping of high-dimensional data to low-dimensional space, resulting in dimensionality-reduced data, effectively removing redundant interference information and extracting core fault features.

[0037] Step 3: Sample set construction Based on the dimensionality-reduced data, training, validation, and test sets are constructed. Stratified sampling is used to partition the samples (a recommended ratio of training:validation:test = 6:2:2) to ensure that the equipment fault types and operating conditions in each set are consistent with the original data, avoiding imbalanced sample distribution that could lead to decreased model generalization ability. The training set is used for model parameter fitting, the validation set for hyperparameter tuning and overfitting monitoring, and the test set for evaluating the model's final detection performance.

[0038] Step 4: Deep Learning Model Training and Prediction Model Construction The constructed deep learning model was trained using training, validation, and test sample sets to obtain a high-precision prediction model. A feedforward neural network was chosen as the deep learning model due to its simple structure, fast inference speed, and suitability for real-time fault detection in electrical equipment. The feedforward neural network consists of two hidden layers. The first hidden layer contains 48 neurons, combined with batch normalization (BN) and L2 regularization: BN accelerates model training convergence and alleviates the gradient vanishing problem, while L2 regularization reduces model complexity by penalizing weight parameters, effectively suppressing overfitting. The second hidden layer contains 24 neurons, combined with batch normalization and Dropout (with a deactivation probability set to 0.3): Dropout randomly deactivates some neurons, breaking redundant dependencies between neurons and further enhancing model robustness. Each neuron in both hidden layers uses the ReLU activation function, which effectively fits the nonlinear characteristics of the data while avoiding gradient saturation.

[0039] The output layer of the feedforward neural network is a single neuron layer, and the output value corresponds to the equipment failure risk coefficient (0~1, the closer to 1, the higher the failure risk). The mean squared error (MSE) is used as the loss function to quantify the difference between the predicted value and the actual value. This loss function is sensitive to small errors and is suitable for the detection of weak early failure signals. The Adam optimizer is used to adjust the network weights. This optimizer combines momentum gradient descent and adaptive learning rate strategies, which has a faster convergence speed, stronger stability, and can accurately optimize model parameters. An early stopping strategy is used during training to prevent overfitting: when the validation set loss does not decrease for 10 consecutive rounds, the model training is stopped, and the model parameters with the minimum validation set loss are retained.

[0040] The specific process of training a prediction model using a sample set includes: First, using the training sample set and validation set, the constructed deep learning model is iteratively trained to obtain intermediate prediction models adapted to the characteristics of different target objects such as transformers, circuit breakers, and capacitors. Based on the intermediate prediction model of each target object, an online learning algorithm is used to fine-tune the model parameters periodically (recommended every 72 hours) using new data from real-time equipment operation, so as to realize the dynamic updating and optimization of the model, adapt to dynamic scenarios such as equipment aging and changes in operating conditions, and finally obtain a prediction model with strong generalization ability and high adaptability.

[0041] Step 5: Fault Detection Based on Predictive Models Based on the trained and optimized prediction model, real-time fault detection is performed on the current primary electrical equipment. The specific process is as follows: real-time collection of multi-dimensional parameters of the current equipment; handling of outliers and missing values ​​according to the preprocessing standards in step 1; and then standardization and principal component dimensionality reduction in step 2, followed by input into the prediction model. The model outputs the equipment fault risk coefficient and fault type assessment results (normal, minor anomaly, severe fault). If the detection result is a minor anomaly, early fault warning information is automatically generated (including suspected fault location, risk level, and recommended maintenance measures). If the detection result is a severe fault, an emergency warning is immediately triggered, and the power operation and maintenance system is linked to push alarm information to the operation and maintenance terminal. At the same time, the operating parameters of the equipment before and after the fault are recorded to provide data support for fault tracing and realize closed-loop fault management throughout the entire process of "detection-early warning-tracing".

[0042] like Figure 2 As shown, based on the above method, the present invention provides a fault detection device for primary electrical equipment, comprising: a data acquisition unit 201 for acquiring parameters of different primary electrical equipment to obtain a corresponding dataset; an analysis unit 202 for performing principal component analysis on the dataset to obtain dimensionality-reduced data; a construction unit 203 for constructing a training sample set and a test sample set based on the dimensionality-reduced data; a training unit 204 for training a deep learning model using the training sample set and the test sample set to obtain a prediction model; and a detection unit 205 for detecting faults in the current primary electrical equipment based on the prediction model.

[0043] It should be understood that all relevant content of each step involved in the above method embodiments can be referenced to the functional description of the corresponding functional module, and will not be repeated here. Furthermore, the use of suffixes such as "module," "component," or "unit" to represent elements is merely for the purpose of illustrative purposes and has no specific meaning in itself. Therefore, "module," "component," or "unit" can be used interchangeably. Terminals can be implemented in various forms. For example, the terminals described in this invention may include mobile terminals such as mobile phones, tablets, laptops, handheld computers, personal digital assistants (PDAs), portable media players (PMPs), navigation devices, wearable devices, smart bracelets, pedometers, etc., as well as fixed terminals such as digital TVs and desktop computers. The following description will use mobile terminals as examples; those skilled in the art will understand that, in addition to elements specifically designed for mobile purposes, the construction according to embodiments of the present invention can also be applied to fixed-type terminals.

[0044] In other embodiments of the present invention, an electronic device 300 is disclosed, such as... Figure 3 As shown, the device may include: one or more processors 301; memory 302; display 303; one or more application programs (not shown); and one or more computer programs 304. These devices can be connected via one or more communication buses 305. The one or more computer programs 304 are stored in the memory 302 and configured to be executed by the one or more processors 301. The one or more computer programs 304 include instructions that can be used to perform actions such as... Figure 1 Each step in the corresponding embodiment.

[0045] Processor 301 can 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. A general-purpose processor can be a microprocessor or any conventional processor.

[0046] The memory 302 can be an internal storage unit of the electronic device 300, such as a hard disk or RAM of the electronic device 300. The memory 302 can also be an external storage device of the electronic device 300, such as a plug-in hard disk, Smart Media Card (SMC), Secure Digital (SD) card, or FlashCard equipped on the electronic device 300. Furthermore, the memory 302 can include both internal and external storage units of the electronic device 300. The memory 302 is used to store computer programs and other programs and data required by the electronic device. The memory 302 can also be used to temporarily store data that has been output or will be output.

[0047] The computer program 304 can be divided into one or more modules / units. The one or more modules / units can be a series of computer program instruction segments that can perform a specific function. The instruction segments are used to describe the execution process of the computer program 304 in the electronic device 300.

[0048] In addition to the above-described structure, those skilled in the art will understand that Figure 3 This is merely an example of electronic device 300 and does not constitute a limitation on electronic device 300. Electronic device 300 may include more or fewer components than shown, or combine certain components, or different components. For example, electronic device may also include input / output devices, network access devices, buses, etc.

[0049] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is merely an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of the functional units and modules are only for easy differentiation and are not intended to limit the scope of protection of this invention. The specific working process of the units and modules in the above system can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0050] Based on the above embodiments, the present invention also discloses a computer-readable storage medium having at least one computer program stored thereon, wherein the computer program, when executed by a processor, implements the methods described in the foregoing embodiments.

[0051] Those skilled in the art will understand that all or part of the steps in the methods of the above embodiments can be implemented by a program instructing a processor, and the program can be stored in a computer-readable storage medium. The storage medium is a non-transitory medium, such as random access memory, read-only memory, flash memory, hard disk, solid-state drive, magnetic tape, floppy disk, optical disk, and any combination thereof. The aforementioned storage medium can be any available medium accessible to a computer or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., digital video disc (DVD)), or a semiconductor medium (e.g., solid-state disk (SSD)).

[0052] The descriptions of the processes or structures corresponding to the above figures each have their own emphasis. For parts of a process or structure that are not described in detail, please refer to the relevant descriptions of other processes or structures.

[0053] In summary, the present invention discloses a method, apparatus, device, medium, and program for detecting faults in primary electrical equipment. It utilizes principal component analysis combined with deep learning technology to construct a dataset by collecting parameters from different power equipment. Principal component analysis is used to reduce the dimensionality of the data to remove interference information and extract core features. Then, a sample set is constructed based on the dimensionality-reduced data to train a deep learning model. Finally, the trained prediction model is used to achieve accurate fault detection of the current power equipment. It can adapt to the dynamic operating conditions of the equipment, has strong anti-signal interference capabilities, provides reliable data, and can accurately capture early fault signals.

[0054] Although the embodiments of the present invention have been described in detail above, it will be apparent to those skilled in the art that various modifications and variations can be made to these embodiments. The above are merely embodiments of the present invention and do not limit the patent scope of the present invention. However, it should be understood that such modifications and variations fall within the scope and spirit of the present invention. Moreover, the present invention described herein may have other embodiments and can be implemented or realized in various ways. All equivalent transformations made based on the description and drawings of the present invention, or direct or indirect applications in related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A method for fault detection in primary electrical equipment, characterized in that, include: Collect parameters from different primary electrical devices to obtain the corresponding datasets; Principal component analysis was performed on the dataset to obtain dimensionality-reduced data; Based on the dimensionality-reduced data, a training sample set and a test sample set are constructed; The deep learning model is trained using the training sample set and the test sample set to obtain the prediction model; The prediction model is used to detect faults in the current primary electrical equipment.

2. The method according to claim 1, characterized in that, Parameters from different primary electrical devices were collected to obtain the corresponding datasets, including: Collect parameters from different primary electrical equipment, preprocess all parameters, and obtain preprocessed data; Correlation analysis is performed on the preprocessed data to obtain the corresponding dataset; The preprocessing includes outlier handling and missing value handling.

3. The method according to claim 1 or 2, characterized in that, Principal component analysis was performed on the dataset to obtain the dimensionality-reduced data, including: The dataset is standardized to obtain standardized data; Calculate the covariance matrix of the standardized data; The covariance matrix is ​​decomposed into eigenvalues ​​to obtain the magnitudes of the eigenvalues ​​of the principal components. Based on the magnitude of the eigenvalues, select the top 5 principal components with the largest eigenvalues; Construct feature vectors based on the five selected principal components; The dataset is projected onto the principal components to obtain the dimensionality-reduced data.

4. The method according to claim 1, characterized in that, The deep learning model is trained using the training and testing sample sets to obtain the prediction model, which includes: Using the training and testing sample sets, the deep learning model is trained to obtain an intermediate prediction model for each target object; The intermediate prediction model for each target object is updated and optimized regularly to obtain the prediction model.

5. The method according to claim 1, characterized in that, The primary electrical equipment includes transformers, circuit breakers, and capacitors; and / or the parameters include waveforms, amplitudes, frequencies, electromagnetic waves, sound waves, and optical signals.

6. The method according to claim 1, characterized in that, The deep learning model is a feedforward neural network; The feedforward neural network consists of two hidden layers. The first hidden layer contains 48 neurons, with batch normalization and L2 regularization, and the second hidden layer contains 24 neurons, with batch normalization and Dropout. The activation function of each neuron in both hidden layers is the ReLU activation function. And / or the output layer of the feedforward neural network is a single neuron layer, the mean squared error is used as the loss function to quantify the difference between the predicted value and the actual value, the Adam optimizer is used to adjust the weights, and an early stopping strategy is used during training to prevent overfitting.

7. A fault detection device for primary electrical equipment, used in the method according to any one of claims 1 to 6, characterized in that, include: The data acquisition unit is used to collect parameters from different primary electrical devices and obtain the corresponding dataset. The analysis unit is used to perform principal component analysis on the dataset to obtain dimensionality-reduced data. A construction unit is used to construct a training sample set and a test sample set based on the dimensionality-reduced data; The training unit is used to train the deep learning model using the training sample set and the test sample set to obtain the prediction model. The detection unit is used to perform fault detection on the current primary electrical equipment based on the prediction model.

8. An electronic device, characterized in that, The device includes a memory and a processor, wherein the memory stores a program that can run on the processor, and when the program is executed by the processor, causes the electronic device to perform the method of any one of claims 1-6.

9. A readable storage medium storing a program, characterized in that, When the program is executed, it implements the method of any one of claims 1-6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the method of any one of claims 1-6.

Images