REAL-TIME CASCADE ADAPTIVE NOTCH FILTER SCHEME FOR SINUSOIDAL PARAMETER-ESTIMATION

In this paper, a real-time cascade adaptive notch filter scheme for si nusoidal parameter estimation is proposed. The second order recursive maximum likelihood algorithm is used to estimate the sinusoidal freque ncies. It is suitable for real-time operations because only 13 multipl ications, 14 additions and 1 division are required in each iteration. After adaptive notch filters have converged, the simplified recursive least square algorithm is then used to compute the amplitudes and phas es of sinusoids quickly. The error surface analysis indicates that our scheme is unimodal and results in guaranteed convergence. Extensive c omputer simulations are also included to demonstrate its performance.