Fault diagnosis method of power transformer based on extreme learning machine

Abstract

The invention relates to a fault diagnosis method of a power transformer based on an extreme learning machine, which can be applied to a transformer monitoring / detection device or system. Fault characteristics are extracted based on data collected by the monitoring / detection device or system, and fault diagnosis module leaning of the extreme leaning machine of the transformer is carried out by selecting state samples of the transformer. The method comprises the steps of dividing the operating state of the transformer; selecting monitoring / detection data comprising the operating state of the transformer as a data source; extracting characteristics of the data source of the transformer, and determining characteristic variables; determining target vector expression manner of the extreme learning machine of the transformer in various operating states; selecting sample data of the transformer in various operating states; determining training sample data and testing the sample data; determining an input layer, a hidden layer, node number of an output layer and an excitation function of the fault diagnosis module of the extreme learning machine of the transformer; and learning and verifying the fault diagnosis module of the extreme learning machine of the transformer.

Description

technical field [0001] The invention relates to a method for fault diagnosis of a power transformer, which belongs to the technical field of fault diagnosis of power equipment. Background technique [0002] Transformer is an important equipment in the power system, and its operating status directly affects the safety level of the system. At present, there are many kinds of monitoring / detection including oil chromatography monitoring / detection (gas content analysis in oil DGA), partial discharge monitoring / detection (pulse current, ultra-high frequency UHF and ultrasonic method, etc.), broadband online monitoring of grounding current, etc. device or system. It is difficult to complete a large amount of monitoring / testing data by manual analysis. It is necessary to find a fast and automatic fault diagnosis method and embed it into the monitoring / testing device or system, so as to detect potential faults and fault types in time, and provide a basis for condition-based maintena...

AI Technical Summary

Problems solved by technology

[0003] Among the existing transformer fault diagnosis methods, the Bayesian network diagnosis method requires a large number of sample data, and the characteristic variable is a discrete variable. However, there is no theoretical basis for the selection of the discrete threshold value, and the discrete process will cause the loss of transformer state information; the support vector machine diagnosis method It is difficult to determine regularization parameters and kernel function parameters. In addition, transformer fault diagnosis is essentially a multi-classification problem, while support vector machine is a binary classification algorithm. Converted to multi-classification, there are problems such as classification overlap and non-classification, more classifiers need to be built, and error accumulation; the neural network diagnosis method needs to solve many network parameters, and it needs to be determined iteratively during the training process, and the search space and calculation amount are very large. Large, it is necessary to select an appropriate learning rate and initial value of input weights to achieve ideal results, and may fall into local optimum

Images