In order to improve the alarm accuracy rate in nuclear power station, this paper adopts Wavelet De noising (WD) method based on statistic characteristics and presents a loose parts alarm method.
针对核电站中检测跌落零件误报警率高的问题，本文对带通滤波后的跌落零件冲击信号应用了基于统计特征的小波去噪方法，并给出了去噪后的跌落零件报警方法。
Translation invariant(TI)wavelet de-noising method is an improved method of wavelet thresholding shrinks; it can suppress Pseudo-Gibbs artifacts effectively.
图像的平移不变量小波去噪法是小波收缩阈值去噪法的改进，能够有效抑制小波阈值去噪过程中出现的伪吉布斯现象。
Finally, the parameter setups for wavelet de-noising function are analyzed in the thesis, and the different results with different parameters have been compared each other for the final conclusions.
最后本文对小波消噪的参数设置问题进行了分析，对不同参数选择后的结果进行了比较后得出了结论。
应用推荐