Abstract:

As the noise in the acceleration signal of the acquired shift condition has a great influence on the accuracy and complexity of the excavation of the shift quality evaluation index, a signal denoising method was proposed based on Empirical Mode Decomposition (EMD) and Wavelet Threshold. Firstly, EMD was used to decompose the original noisy signal into finite Intrinsic mode function (imf) components. Secondly, the high frequency components were denoised by wavelet threshold. Then the high frequency denoising components and low frequency components were reconstructed by EMD to obtain the denoising signal. Finally, the acceleration signal of a certain shift condition was subjected to denoising test. The test results show that compared with the traditional EMD denoising and wavelet threshold denoising, the denoising method based on EMD and wavelet threshold can better preserve the original signal characteristic form and the denoising effect is more obvious. This method provides a feasible idea for the denoising of the shift acceleration signal, which lays foundations for the subsequent drviability evaluation.

摘要：

针对采集的换挡工况加速度信号中存在的噪声对换挡品质评价指标值提取的准确性和复杂性产生较大影响的问题，提出一种基于经验模态分解（EMD）和小波阈值方法结合的信号降噪方法。首先利用EMD将含噪的原始信号分解为有限个本征模态函数（imf）分量，然后对高频imf分量进行小波阈值降噪，并将降噪后高频分量和低频分量利用EMD进行重构得到降噪后的信号。最后对某一换挡工况的加速度信号进行降噪试验。试验结果表明，在换挡加速度信号降噪方面，基于EMD和小波阈值的降噪方法与传统的EMD分解降噪、小波阈值降噪相比，能够更好地保留原始信号特征形态，降噪效果更明显。该方法为换挡加速度信号的降噪处理提供了一种可行的思路，为后续整车驾驶性评价创造了条件。

关键词: 振动与波；换挡工况；经验模态分解, 小波阈值降噪；本征模态函数