小波分析在1/f噪声分析中的应用

Wavelet analysis in 1/f noise analysis application

提出一种基于正交基展开方法的脉冲响应辨识方法——正交尺度变换法。这种方法以正交尺度函数展开的形式表示脉冲响应函数,对信号进行正交尺度变换,并使用最小二乘类参数辨识方法估计参数。这里提出的方法只需辨识g(t)在最大的尺度下的正交小波变换系数,使得辨识的计算量大大减少。仿真结果表明,在辨识所用的数据长度相同的前提下,该辨识方法不仅具有高精度,而且具有强跟踪参数变化能力。

The method of pulse response identification based on orthogonal expansion was proposed. This method expressed the pulse response function in the form of orthogonal scaling function expansion, carried on the orthogonal scaling transform to the signal, and used the least-squares parameter identification methods to estimate parameter of class. The method presented in this paper identified g（t） only in the largest scales of the orthogonal wavelet transform coefficients, which makes identification of the computation greatly reduced. The simulation result indicated that this identification method not only has a higher precision, but also has a stronger ability to change tracking parameters under the premise of the same data length.