微弱信号检测的一种改进小波分析方法

An Improved Method of Wavelet Analysis for Weak Signal Detection

针对小波变换去噪时对微弱信号不敏感,特别是在信号分类检测时检测效果不理想的问题,提出了采用小波熵应用于信号检测的技术.该技术的主要原理是利用小波分析矩阵相应稀疏程度去抑制信号的无关成份,其中小波系数熵确定噪声阈值,小波相关系数熵确定边界轮廓,实现信号准确定位.最后通过数值试验证明其效果好于常规小波变换模极大值的信号检测法,证明了该方法检测微弱信号边缘特性更切实有效.

According to wavelet transform to noise,the weak signal for a small value or not sensitive signal,especially for the problem of unsatisfactory results in signal classification test,puts forward the wavelet entropy used in signal detection technology.The technology is by using the wavelet matrix to suppress the sparse degree not components,including wavelet coefficients are used to determine the noise threshold entropy,wavelet entropy are used to determine the correlation coefficient boundary contours,signal a method of accurate positioning.Finally by numerical experiments prove good effect in the conventional wavelet transform modulus maximum signal method,this method proves a weak signal detection edge characteristic more practical and effective.