Search

CN-121997123-A - Fault detection method, device, apparatus, medium, and program for electric primary equipment

CN121997123ACN 121997123 ACN121997123 ACN 121997123ACN-121997123-A

Abstract

The invention provides a fault detection method, device, equipment, medium and program for electrical primary equipment, which are used for acquiring parameters of different electrical primary equipment to obtain corresponding data sets, carrying out principal component analysis on the data sets to obtain dimension-reduced data, constructing a training sample set and a test sample set based on the dimension-reduced data, training a deep learning model by utilizing the training sample set and the test sample set to obtain a prediction model, and carrying out fault detection on current electrical primary equipment based on the prediction model. The method has the advantages that the data set is built by collecting parameters of different power equipment, the data dimension reduction is realized by means of principal component analysis to remove interference information and refine core characteristics, the sample set training deep learning model is built based on the dimension reduction data, the accurate fault detection of the current power equipment is finally realized by utilizing the prediction model obtained by training, the dynamic working condition of the equipment can be adapted, the signal interference resistance is high, the data is reliable, and early fault signals can be accurately captured.

Inventors

  • HUANG XUEQIONG
  • WANG JUYAN
  • LI PENG
  • WU JINGJING

Assignees

  • 福州优能电子科技有限公司

Dates

Publication Date
20260508
Application Date
20251231

Claims (10)

  1. 1. A method of detecting a failure of an electrical primary device, comprising: acquiring parameters of different electrical primary equipment to obtain corresponding data sets; Performing principal component analysis on the data set to obtain dimension-reduced data; Constructing a training sample set and a testing sample set based on the dimension reduced data; 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 electric primary equipment based on the prediction model.
  2. 2. The method of claim 1, wherein collecting parameters of different electrical primary devices to obtain corresponding data sets comprises: Collecting parameters of different electrical primary equipment, and preprocessing all the parameters to obtain preprocessed data; Performing correlation analysis on the preprocessed data to obtain a corresponding data set; the preprocessing includes outlier processing and missing value processing.
  3. 3. The method of claim 1 or 2, wherein performing a principal component analysis on the dataset to obtain reduced-dimension data comprises: carrying out standardization processing on the data set to obtain standardized data; Calculating a covariance matrix of the standardized data; Performing eigenvalue decomposition on the covariance matrix to obtain the eigenvalue of the main component; selecting the first 5 main components with large characteristic values according to the characteristic values; Constructing a feature vector according to the selected 5 principal components; and projecting the data set onto the principal component to obtain the dimension-reduced data.
  4. 4. The method of claim 1, wherein training the constructed deep learning model using the training sample set and the test sample set to obtain the predictive model comprises: Training the constructed deep learning model by using the training sample set and the test sample set to obtain an intermediate prediction model of each target object; and (5) updating and optimizing periodically based on the intermediate prediction model of each target object to obtain a prediction model.
  5. 5. The method of claim 1, wherein the electrical primary device comprises a transformer, a circuit breaker, and a capacitor, and/or the parameter comprises a waveform, an amplitude, a frequency, an electromagnetic wave, an acoustic wave, and an optical signal.
  6. 6. The method of claim 1, wherein the deep learning model is a feed-forward neural network; The feedforward neural network consists of two hidden layers, wherein the first hidden layer comprises 48 neurons and is matched with batch normalization and L2 regularization, the second hidden layer comprises 24 neurons and is matched with batch normalization and Dropout, and each neuron activation function of the two hidden layers is a ReLU activation function; And/or the output layer of the feedforward neural network is a single neuron layer, the difference between the predicted value and the actual value is quantized by taking the mean square error as a loss function, the weight is adjusted by using an Adam optimizer, and an early-stop strategy is adopted in the training process to prevent overfitting.
  7. 7. An electrical primary equipment failure detection apparatus for use in the method of any one of claims 1 to 6, comprising: the acquisition unit is used for acquiring parameters of different electrical primary equipment to obtain corresponding data sets; The analysis unit is used for carrying out principal component analysis on the data set to obtain dimension-reduced data; the construction unit is used for constructing a training sample set and a testing sample set based on the dimension-reduced data; the training unit is used for training the constructed deep learning model by utilizing the training sample set and the test sample set to obtain a prediction model; and the detection unit is used for detecting faults of the current electric primary equipment based on the prediction model.
  8. 8. An electronic device comprising a memory and a processor, the memory having stored thereon a program executable on the processor, which when executed by the processor, causes the electronic device to implement the method of any of claims 1-6.
  9. 9. A readable storage medium having a program stored therein, characterized in that the program, when executed, implements the method of any one of claims 1-6.
  10. 10. A computer program product comprising a computer program, characterized in that the computer program, when executed by a processor, implements the method of any of claims 1-6.

Description

Fault detection method, device, apparatus, medium, and program for electric primary equipment Technical Field The present invention relates to the field of electrical detection, and in particular, to a method, apparatus, device, medium, and program for detecting faults of an electrical primary device. Background The electric primary equipment refers to physical devices (such as a generator, a transformer, a circuit breaker, a power cable and the like) directly participating in the production, transmission, distribution and consumption of electric energy, and is used as a core component of a power system, and the stability of the running state of the electric primary equipment directly relates to the safe and reliable power supply of a power grid. Therefore, the method and the device perform accurate and real-time fault detection on the electric primary equipment, timely find potential fault hidden dangers and take intervention measures, and are key links for guaranteeing stable operation of the electric power system, reducing operation and maintenance cost and avoiding large-area power failure accidents. Along with the development of the power system in the intelligent and large-scale directions, higher requirements are put forward on the precision, adaptability and reliability of the fault detection technology of the electric primary equipment. However, the existing electrical primary equipment detection technology still has a plurality of key defects, and is difficult to meet actual operation demands, and the specific characteristics are that firstly, detection parameters are fixed preset values, dynamic operation conditions of equipment under different loads, environment temperatures and voltage levels cannot be dynamically adapted, so that detection standards are not matched with actual operation conditions, further fault detection accuracy is remarkably reduced, potential faults caused by working condition changes cannot be accurately identified, secondly, multiple types of detection signals such as electromagnetic waves, sound waves and electric signals are involved in the detection process, various signals are prone to cross interference, the identification degree of effective fault signals is reduced, the fault misjudgment rate is greatly increased, a large amount of operation and maintenance resources are wasted, normal operation equipment is stopped due to misjudgment, and power supply stability is affected, thirdly, the existing detection device lacks an effective real-time calibration mechanism, is affected by factors such as environment erosion and part aging in the long-term operation process, the accuracy drift phenomenon is prone to occur, the reliability of collected data is reduced, fault judgment is not accurately carried out based on unreliable data, further misjudgment and leakage risk is further caused, fourth, the electrical primary equipment is prone to failure is greatly attenuated, once the existing fault signals are difficult to be well attenuated, and the existing fault signals are difficult to be well attenuated in the process is difficult to be greatly attenuated, and the fault is difficult to be well lost, and the fault is caused to be partially lost. Accordingly, there is a need for an electrical primary device fault detection method, apparatus, device, medium, and program that ameliorates the foregoing problems. Disclosure of Invention The invention aims to provide a fault detection method, device, equipment, medium and program for electrical primary equipment, which can adapt to the dynamic working condition of the equipment, has strong signal interference resistance, reliable data and can accurately capture early fault signals. In a first aspect, the present invention provides a method for detecting a failure of an electrical primary device, including: acquiring parameters of different electrical primary equipment to obtain corresponding data sets; Performing principal component analysis on the data set to obtain dimension-reduced data; constructing a training sample set and a testing sample set based on the dimension reduced data; training the constructed deep learning model by using a training sample set and a test sample set to obtain a prediction model; and performing fault detection on the current electrical primary equipment based on the prediction model. Optionally, collecting parameters of different electrical primary devices, obtaining a corresponding data set includes: Collecting parameters of different electrical primary equipment, and preprocessing all the parameters to obtain preprocessed data; carrying out correlation analysis on the preprocessed data to obtain a corresponding data set; the preprocessing includes outlier processing and missing value processing. Optionally, performing principal component analysis on the data set to obtain the dimension-reduced data includes: Carrying out standardization processing on the data set to obtain standardized data; c