平稳小波自适应去噪用于曲轴瞬时角加速度测量

针对曲轴角加速度测量中的噪声放大问题,构造了基于平稳小波变换的自适应去噪方案。该方案以小波系数能量熵最小为原则选择小波函数,根据各层小波方差的变化速率确定所需的分解层数,并由小波系数幅值分布状态决定阈值的生成准则。对比了平稳小波自适应去噪与常用的6阶Butterworth滤波器对曲轴角加速度噪声的抑制能力。结果表明,当柴油机状态正常时,两种方法去噪得到的结果相当;当柴油机高速轻载运行且有连续失火故障时,只有平稳小波自适应去噪能保留由失火引起的局部高频振荡信号。

Measurement of Instantaneous Angular Acceleration of Crankshaft Using Adaptive Stationary Wavelet Denoising

An adaptive stationary wavelet denoising method is presented for measur ing instantaneous angular acceleration of a crankshaft. The wavelet function is selected by minimizing the energy entropy of wavelet coefficients, the decomposi tion level is determined by checking the variation rate of the wavelet variance,and the method to calculate the threshold of each level is chosen according to the distribution of wavelet coefficients. The proposed denoising method is compa red with the 6 order Butterworth filterer, which is widely used in the instanta n eous angular speed measurement of crankshafts. The results show that if the diesel engine is in normal conditions, two methods have similar performance. While it runs at high speeds, under low loads, and with continuous misfires, only the a daptive stationary wavelet denoising method can preserve local high frequency os cillation signals caused by misfires.