Power transformer fault diagnosis method based on relative transformation and kernel entropy component analysis

Abstract

The invention discloses a power transformer fault diagnosis method based on relative transformation and kernel entropy component analysis, which adopts the relative transformation kernel entropy component analysis method to extract features, constructs a training sample set and a test sample set of a kernel limit learning machine, and then trains a kernel limit learning machine classifier to forma final RTKECA-KELM diagnostic model; optimization of RTKECA by Adaptive Quantum Particle Swarm Optimization-KELM diagnoses the model parameters, obtains the optimal model parameter combination, and saves the optimal RTKECA-KELM diagnostic model; entering test samples or samples to be diagnosed into the trained optimal RTKECA-KELM diagnostic model. The invention not only can utilize the advantagesof nonlinear amplification and noise suppression of RT, but also can give play to the advantages of KECA non-linear feature extraction, has good feature extraction effect, has higher fault diagnosisaccuracy, and effectively improves the fault diagnosis accuracy of the model.

Description

technical field [0001] The invention relates to the technical field of transformer fault diagnosis, in particular to a power transformer fault diagnosis method based on relative transformation and nuclear entropy component analysis. Background technique [0002] Ensuring the safe and stable operation of power transformers is one of the keys to improving the safety level of the entire power system. The application of fault diagnosis technology is an important means to ensure the safety of power transformers. Research on power transformer fault diagnosis methods to timely and accurately judge its fault status has important practical significance. Dissolved gas analysis in oil (DGA, dissolved gas analysis) can provide an important basis for transformer fault diagnosis. In the past ten years, with the development of artificial intelligence, machine learning, data mining and other technologies, artificial intelligence diagnosis methods based on DGA data, such as support Vector m...

AI Technical Summary

Problems solved by technology

For example, GA has a long iterative convergence time in the calculation process, and PSO is easily affected by the local particle optimum, and "premature phenomenon" appears, etc.

Images