作者: A. Glowacz
关键词: Process (computing) 、 DC motor 、 Signal processing 、 Bayes classifier 、 Electronic engineering 、 Electric motor 、 Induction motor 、 Engineering 、 Fault (power engineering) 、 Rotor (electric)
摘要: In this paper, a non-invasive method of early fault diagnostics electric motors was proposed. This uses acoustic signals generated by motors. Essential features were extracted from A plan study Researches carried out for faultless induction motor, motor with one faulty rotor bar, two bars and flawless Direct Current, Current shorted coils. methods signal processing: log area ratio coefficients, Multiple classification, Nearest Neighbor classifier the Bayes classifier. pattern creation process using 40 samples sound. identification 130 five-second test used. The proposed approach will also reduce costs maintenance number in industry.