针对传统小波脊线法在暂态信号检测中存在的问题,提出了改进算法。新方法引入小波包变换,扩展了小波脊线法的应用范围,并对算法本身进行了若干改进,基本消除了边界效应,提高了算法进行时频分析的实时性和精度。通过对仿真信号的分析,验证了改进方法的优越性,表明改进算法能有效地消除噪声干扰,提取暂态信号的瞬时频率和瞬时幅值,获得信号完整的、精确的时频分布。
Aiming at the problems of the traditional wavelet-ridge method applied in transient signal's detection, this paper puts forward an improved method. By introducing the wavelet-packet transformation, the application area of the wavelet-ridge method is extended. And by improving the method, the boundary effect is mostly eliminated, and the real-time performance and precision are increased. The analysis to the stimulation signal verifies the superiority of the improved method. It also indicates that the improved method can efficiently eliminate the disturbance of noises, abstract the instantaneous frequency and amplitude of the transient signal and get the complete and accurate time-frequency distribution of signals.