摘要
对于存在偏心故障的感应电动机,检测其振动信号,并应用连续小波变换进行信号分析,判断是否出现碰摩故障。通过增加Morlet小波函数的时域波形波峰数,减小Morlet小波函数的频域带宽,从而减轻采用Morlet小波函数进行连续小波变换时的频域混叠现象。对于偏心故障及偏心和碰摩同时存在的复合故障信号,应用连续小波变换提取中心频率为25、75和100Hz的3个子频段信号。引入子频段提取信号的能量表达式,计算得到两种故障形式子频段信号能量值并进行对比。将中心频率25、75Hz这2个子频段提取信号能量值作为区分两种不同故障形式的特征量。通过实验采集三相感应电动机偏心,偏心+轻微碰摩,偏心+严重碰摩3组故障信号并进行信号分析,以验证该方法的有效性。
The purpose of this paper is to detect vibration signals of induction motors with eccentricity, analyse them by continuous wavelet transfrom（CWT）, and determine whether rub fault happens. Decrease bandwidth of Morlet wavelet through increasing it＇s time-domain wave crests can alleviate frequency aliasing in CWT. Pick up sub-frequency signals of eccentricity and mixed faults of eccentrity plus rub faults using CWT. Three sub-frequency bandwidths are centring around 25, 75 and 100 Hz. Introduce a formula to express energy of sub-frenqucy signals. Compute then obtain energy magnitudes of two different faults. Contrast the magnitudes of these energies. Regard energies of 25 and 75 Hz sub-frenquency signals as characteristics to distinguish these two faults. Finally gather three fault signals of an three-phase induction motor, which are eccentricity, mixed fault of eccentricy plus mild rub and mixed fault of eccentricy plus moderate rub through experiments. Results of experiments and analysis show that techniques in this paper are effective.
出处
《中国电机工程学报》
EI
CSCD
北大核心
2006年第8期159-162,共4页
Proceedings of the CSEE
