|
标题：电机转子重复脉冲故障诊断小波去噪研究 |
|
作者：李天辉，唐明，潘瑾，等 |
|
| 2018年第11期 访问次数：235次 |
|
摘要：基于重复脉冲技术对电机转子绕组进行故障诊断，具有检测效率高、预警性强以及定位准确程度高等优点。基于互相关系数在coif4、sym6以及db6等母小波中选取最优母小波，提出综合参数选取分解层数。对去噪后的特征波形利用阈值法，能量法及AIC法等3种方法分别计算起始时刻，并比较去噪前后起始时刻的计算精度。对比结果表明：小波变换去噪后特征波形更平滑，起始时刻提取更准确，定位效果可限定在一匝线圈中。
关键词：小波变换；起始时刻；特征波形；小波去噪
Abstract: Fault diagnosis of motor rotor winding based on repetitive pulse technology has the advantages of high detection efficiency, strong early warning and high accuracy of location. Based on the correlation number, the optimal mother wavelet is selected in coif4, sym6 and db6 mother wavelets, and the synthetic parameters is proposed to selection the decomposition layers. Using the threshold method, energy method and AIC method to calculate the staring time of the denoised feature waveforms, and compared the calculation accuracy of the starting time before and after denoising. The comparison results show that the characteristic waveform of wavelet transform is smoother, the start time is more accurate, and the localization effect can be limited to one turn coil.
Key words: Wavelet transform; starting time ; characteristic waveform; wavelet denoising |
|
| |
|
