The research on LMS adaptive filter based on multi-scale domain

This paper researches on designing LMS adaptive filter by using multi-scale wavelet theory. With presenting the structure of linear filter based on Orthogonal wavelet, analyzing theory of adaptive filter algorithm based on multi scale orthogonal wavelet and integrating LMS algorithm with multi-scale wavelet transform, a new adaptive filter algorithm is proposed. According to the new algorithm, autocorrelation matrix condition number is decreased, the contradictions between convergence rate, precision and step factor rare are overcome. Then, the better convergence rate and stability can be obtained. At last, The new filter algorithm is applied in signals processing for servo control system, the simulation results show that the new filter avoids slow convergence and big imbalance in traditional adaptive filter system, and the new algorithm is effective.