摘要
为了获取高信噪比的柴油机气缸压力信号,采用非线性离散序列的小波变换(DTW T)对柴油机气缸压力信号进行降噪.由于降低压力信号采集循环数目,简要分析了气缸压力信号中的高频噪声和高斯白噪声特征,先用D onoho的阈值降噪法去除高斯白噪声,降噪效果明显优于低通滤波和9点平滑法;再用多分辨率分析法,分析不同频率段上的噪声幅值,用小波尺度系数反变换来去除高频噪声,恢复柴油机气缸压力信号.结果表明:这种方法能够有效地去除气缸压力信号的噪声.
To obtain high signal-noise ratio signal of cylinder pressure in diesel engine, de-noising processing is presented by using nonlinear discrete time wavelet transform （DTWT）. After the article analyzing high frequency noise and Gauss white noise existing in the cylinder pressure signal, the Donoho＇s soft-threshold de-noising method is firstly used to wipe off the Gauss white noise, the effect of which is superior to that of low pass filter and that of nine-dot smoothing way. Then based on muhiresolution analysis theory, the noise amplitudes of different frequency segments are calculated. Finally high frequency noise is reduced by the wavelet anti-transform algorithm and the cylinder pressure signal can be restored. Result shows that the presented method can denoise the cylinder pressure signal effectively.
出处
《测试技术学报》
2008年第2期140-144,共5页
Journal of Test and Measurement Technology
基金
国防重点预研基金资助项目(40402020105)
关键词
噪声
气缸压力
小波变换
柴油机
noise
cylinder pressure
wavelet transform
diesel engine
