Transformer Vibration Signal Feature Extraction Based on Compressed Sensing and Wavelet Packet

Abstract Transformers are important equipment in the power system. Aiming at collection, transmission, storage, and processing of massive high-dimensional vibration data process transformer online monitoring, this paper proposes a signal feature extraction method based on compressed sensing wavelet packets. Firstly, PartHadamard measurement matrix is used to compress signal, then characteristics extracted packet decomposition normalized information entropy. The proposed completes measured transformer, three classification algorithms applied carry out simulation experiments. results show that can effectively extract reflecting operating state while greatly reducing dimension data, which provides reference for monitoring vibration.