Abnormal Signal Detection in High-speed and Real-time by usingInnovative Parasitic Discrete Wavelet Transform

到目前为止，作为分析信号的方法，已经使用了许多频率分析方法。重点介绍了小波变换在非平稳信号分析中的有效性。本文的研究目的是对作者提出的可变滤波器带离散小波变换（VFB-DWT）进行以下两点改进：（1）通过对VFB-DWT的分析过程，解决信号在特定频带上的能量损失问题，使信号功率降低;（2）构造VFB-DWT阈值确定的辅助方法。最后，通过本研究提高了分析精度，并通过在胎儿心脏信号分析和车内道路噪声分析中的应用验证了其有效性。

Until now, as the method of analyzing signal, a lot of frequency analysis methods have been used. The wavelet transform which has effectiveness in analysis for non-stationary signals is focused. This research aims to improve the following 2 points in Variable Filter Band Discrete Wavelet Transform (VFB-DWT) which is proposed by author; (1) solving the energy loss problem which makes reducing of signal power atthe specific frequency band in an analytical signal through the analysis process of the VFB-DWT , (2) constructing the assistance method to determine the appropriatethreshold value of VFB-DWT. Finally, improvement of analysis accuracy has been achieved through this research, and its effectiveness has been confirmed through the application in fetal cardiac signals analysis and in-vehicle road noise analysis.