摘要
近几年来小波变换在信号处理尤其在电力系统故障信号的分解、消噪、重构以及故障特征提取等诸多方面得到了广泛的应用 ,基于小波变换的各种算法不断出现。文 [8]提出了使用二进小波变换提取信号边缘特征 ,根据信号特征点的值和导数值用三次埃米特多项式进行插值重构。本文分析了文 [8]存在的两个问题 ,并针对这两个问题进行改进 ,即在二进小波变换和插值重构时使用同一种函数——三角样条小波函数 ,这样才能体现出信号处理的本质。本文作者提出的三角样条小波正好同时具有作为小波函数和插值函数双重作用 ,大大提高算法的效果。本文将它应用到电机故障信号的重构过程中 ,并就信噪比和相对误差与 Mallat算法和文[8]算法进行了比较 。
Recently,the wavelet transforms are widely used in signal processing,especially in the processing of fault signals in power system,such as the decompression,noise reduction,reconstruction and extraction,at the same time,various algorithms based on wavelet transform are unceasingly appearing.A reconstruction algorithm by cubic Hermite spline function interpolation,which is used to recover a signal from its dyadic wavelet transform extreme,is proposed in reference,and the two problems existing in reference are analyzed in this paper and some improvements are made in which the same function i.e.,the trigonometric spline function,is used in dyadic wavelet transform and interpolating reconstruction,in this way the essence of signal processing is incarnated.The trigonometric spline function proposed by the author can just play the double roles of wavelet function and interpolation function,using which the trigonometric spline function the effectiveness of the algorithm can be improved,for example.Applying the trigonometric spline function to the reconstruction of fault signal of electric motor,the obtained signal to noise ratio and relative error are compared with these from Mallat algorithm and the algorithm in reference,the comparison results show that with this method the high signal to noise ratio and low relative errors in signal reconstruction are realizable.
出处
《电网技术》
EI
CSCD
北大核心
2002年第4期9-12,15,共5页
Power System Technology
基金
国家自然科学基金 (50 0 770 0 8)
广东省自然科学基金(980 6 0 8)
关键词
电机故障
信号重构
二进小波变换
三角样条插值
signal reconstruction
dyadic wavelet trans form
trigonometric spline wavelet
spline function interpolation
