Search

CN-121997191-A - Fault detection method, device, equipment, medium and program for electric secondary equipment

CN121997191ACN 121997191 ACN121997191 ACN 121997191ACN-121997191-A

Abstract

The invention provides a fault detection method, device, equipment, medium and program for electric secondary equipment, which are used for acquiring running state information of the electric secondary equipment in a power grid to obtain acquired data, training a big data model by utilizing the acquired data to obtain a trained model, carrying out fault detection on the electric secondary equipment in the current power grid according to the trained model to obtain a detection result, determining a fault processing scheme according to the detection result, and carrying out maintenance dispatch. The factor analysis is combined with the big data technology, a fault detection model which is dynamically adaptive, high in anti-interference capability, accurate and efficient is constructed, the fault detection of the secondary equipment in a full period and real-time is realized, the limitation of the traditional detection method is effectively overcome, potential correlation among the operation parameters of the equipment is excavated in a data driving mode, early fault signals are accurately captured, parameter drift of the equipment after long-term operation is adapted, and reliable guarantee is provided for safe and stable operation of the secondary equipment of the power grid.

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 for detecting a fault in an electrical secondary device, comprising: Acquiring operation state information of electrical secondary equipment in a power grid to obtain acquired data; Training a big data model by using the acquired data to obtain a trained model; performing fault detection on electrical secondary equipment in the current power grid according to the trained model to obtain a detection result; and determining a fault processing scheme according to the detection result, and performing maintenance dispatch.
  2. 2. The method as recited in claim 1, further comprising: preprocessing and dimension reduction processing are carried out on the acquired data; The preprocessing comprises outlier processing, missing value processing and standardization processing, wherein the outlier processing adopts a 3 sigma rule to remove extreme deviation data, the missing value processing adopts an interpolation method to supplement complete data, the dimension reduction processing comprises the steps of calculating a correlation coefficient matrix of the standardized data, judging data suitability, carrying out characteristic value decomposition on the correlation coefficient matrix to obtain the characteristic value of a common factor, screening effective common factors according to the principle that the characteristic value is greater than 1, carrying out orthogonal rotation on the common factors to enable factor meanings to be clearer, constructing factor vectors according to the screened common factors, and projecting a data set onto the common factors to reduce dimension.
  3. 3. The method of claim 1, wherein the big data model is a big data driven gradient lifting tree model, the model is constructed based on XGBoost frames and comprises a feature screening layer, a gradient lifting layer and an output layer, the feature screening layer filters invalid features by a variance threshold method, the gradient lifting layer is composed of 3 lifting units, the first lifting unit comprises 64 decision tree nodes and L1 regularization inhibition overfitting, the second lifting unit comprises 32 decision tree nodes and L2 regularization optimization weights, the third lifting unit comprises 16 decision tree nodes and cross validation optimization parameters, and the output of each lifting unit is summarized by a weighted fusion mode.
  4. 4. The method of claim 3, wherein the output layer of the big data model is a multi-class output unit, a cross entropy loss function is adopted to quantify the difference between the predicted value and the actual value, a parallel gradient descent optimizer is utilized to adjust model parameters, the high-data-rate efficient training requirement is adapted, and an early-stop strategy is adopted to prevent overfitting in the training process.
  5. 5. The method of claim 1, wherein training the big data model by using the collected data comprises dividing the collected data into a sample set and a test set, training the big data model to obtain an intermediate prediction model of each electrical secondary equipment type, and performing incremental updating and parameter optimization by combining real-time collected equipment operation data based on the intermediate prediction model of each target equipment to obtain prediction models adapting to different secondary equipment working conditions.
  6. 6. The method of claim 1, wherein the electrical secondary device comprises a relay, a transformer, a measurement and control device, a protection device, and a fault recorder, and/or the collected data comprises secondary loop voltage, secondary loop current, switching state, message signal, action timing, insulation resistance, and signal amplitude.
  7. 7. An electrical secondary device fault detection apparatus for use in the method of any one of claims 1-6, comprising: the acquisition unit is used for acquiring the running state information of the electric secondary equipment in the power grid to obtain acquired data; The training unit is used for training the big data model by utilizing the acquired data to obtain a trained model; the detection unit is used for carrying out fault detection on the electric secondary equipment in the current power grid according to the trained model to obtain a detection result; and the processing unit is used for determining a fault processing scheme according to the detection result and carrying out maintenance dispatch.
  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, equipment, medium and program for electric secondary equipment Technical Field The present invention relates to the field of electrical detection, and in particular, to a method, an apparatus, a device, a medium, and a program for detecting a fault of an electrical secondary device. Background The electric secondary equipment is used as a neural center and a sensing terminal of the power grid system and bears core functions such as monitoring of the running state of the power grid, transmission of control instructions, rapid response of faults and the like. The running stability of the system directly determines the power supply reliability, the electric energy quality and the operation and maintenance safety of the power grid, and once the system fails, the problems of protection misoperation/refusal, data transmission interruption, abnormal shutdown of equipment and the like can be caused, if the system is light, the local power supply interruption is caused, and if the system is heavy, the system is expanded into systematic power grid accidents, and great economic loss and social influence are caused. With the deep promotion of smart power grid construction, the power grid scale is continuously enlarged, the equipment types are increasingly diversified, the operation working conditions of secondary equipment are increasingly complex, the fault mode of the secondary equipment also shows the characteristics of concealment, diversity, conductivity and the like, and higher requirements are provided for timeliness and accuracy of fault detection. The traditional fault detection of the electrical secondary equipment mainly relies on modes such as manual inspection, periodic verification, threshold value alarm and the like, and has obvious limitation in practical application: The traditional detection is highly subjective, depends on professional experience of operation and maintenance personnel, judges the state of the equipment by means of on-site observation, instrument measurement and the like, is difficult to quantitatively evaluate the operation trend of the equipment, is easily influenced by personal experience difference, and causes high failure omission rate and high misjudgment rate; the anti-interference capability is weak, and complex working conditions are difficult to cope with, namely, a large number of interference factors such as electromagnetic interference, packet loss, parameter fluctuation and the like exist in a secondary circuit, the traditional detection method lacks an effective interference filtering mechanism, and is easy to misjudge an interference signal as a fault signal or mask real fault characteristics due to interference, so that the detection accuracy is influenced; The problems of early performance degradation, potential hidden danger and the like of equipment cannot be found in time, and the problems are often treated after the fault dominance, so that the best maintenance opportunity is missed; the adaptability is insufficient, and the parameter drift of the equipment is difficult to cope with, namely, after the secondary equipment is operated for a long time, the parameter drift can occur due to factors such as aging of components, environmental change and the like, the threshold value and the judgment standard of the traditional detection method are relatively fixed, the change of the operation state of the equipment cannot be dynamically adapted, and the detection precision is obviously reduced after the long-time operation. Accordingly, there is a need for an electrical secondary 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 electric secondary equipment, which can detect faults of the electric secondary equipment in a full period and in real time. In a first aspect, the present invention provides a method for detecting a fault of an electrical secondary device, including: Acquiring operation state information of electrical secondary equipment in a power grid to obtain acquired data; training the big data model by using the acquired data to obtain a trained model; performing fault detection on electrical secondary equipment in the current power grid according to the trained model to obtain a detection result; and determining a fault processing scheme according to the detection result, and performing maintenance dispatch. Optionally, the method further comprises: Preprocessing and dimension reduction processing are carried out on the acquired data; The preprocessing comprises outlier processing, missing value processing and standardization processing, wherein the outlier processing adopts a3 sigma rule to remove extreme deviation data, the missing value processing adopts an interpolation method to supplement complete data, the dimension reduction processing