An adaptive filtering technique with self-adaptive PSO for estimation of non-stationary signals

A hybrid filter using Taylor
^K
Kalman modelling with self-adaptive PSO tuning is presented in this paper for accurately estimating the parameters and harmonic components of a time varying sinusoidal signal in a noisy environment. Second or higher order Taylor expansion modelling of the changing envelope of the power signal and Kalman Filter algorithm for estimating the parameters of the signal model leads us to the design of robust filters referred to as Taylor
^K
Kalman Filters. The proposed algorithm is linear, has less computational complexity and is easier to be implemented on a hardware platform. The use of self-adaptive PSO tune the filter parameters makes the proposed algorithm sensitive to any change in the signal dynamics, immune to noise and harmonic contaminations.