基于双尺度连续小波变换的二相编码信号识别

针对信号和噪声小波变换的模极大值在不同尺度上表现出截然不同的性质,提出一种利用粗细定位相结合的思想对二相编码信号奇异点实现精确定位的方法。该方法采用大尺度上的模极值点进行粗略定位奇异点出现的范围,并据此去除小尺度上的伪极值点,用小尺度上的模极值点精确定位奇异点出现的时刻。实验仿真比较了几种不同小波基函数下的算法性能,结果表明,该方法在采用db3小波基下对二相编码信号突变点的识别具有很好的效果。

Recognition of two phase-code signals based on double-scale CWT

The wavelet transform's modular maximum of signal is contrary to that of noise on different scales,so the singularity of two phase-code signal can be found accurately by the combination of rough orientation and accurate orientation.The modular maximum of a big scale is used to find the range of the singularity,then the false modular maximum is eliminated.In the same way,the exact point of the singularity is found on a small scale.Simulation tests are made to compare the arithmetic performances of different wavelet bases,and the results indicate that this method is significant to the recognition of the singularity of two phase-code signals under db3 wavelet base.