基于声学指纹的换流变状态智能监测及诊断系统开发

Development of Intelligent Monitoring and Diagnosis System for Converter Transformer′s Status Based on Acoustic Fingerprint

换流变运行时的不同声音特征能反映其运行状态,通过对换流变声音进行梅尔倒谱系统MFCC特征提取,研究基于门控循环单元GRU的人工智能算法,开发一套基于声学指纹的换流变状态智能监测及诊断系统。

The characteristics of different sounds generated during the operation of converter transformer can reflect its operating state.Through the MEC feature extraction of Meier inverting spectrum system for commutated variable sound,the artificial intelligence algorithm based on gated cycle unit GRU is studied,and a set of intelligent detection and diagnosis system of converter transformer status based on acoustic fingerprint is developed.