On nonlinear estimation in presence of non-Gaussian Noise

An algorithm for optimal estimation in presence of non-Gaussian observation noise is presented. The algorithm, based on Bayes´ recursion formula is implemented numerically. The filter is shown to be superior to the Kalman Filter when applied to the same system. It has been shown that the steady state estimation error is zeros. The Algorithm is a potential technique for analyzing transients in the automobile electrical environoment.