Two Stage Kalman Estimators with Probabilistically Weighted Average

Abstract

With spherical coordinate, the adaptive estimation using multiple model filtering isenhanced in this paper. The enhancement is achieved by using just two depended parallelKalman filters, instead of multiple models, with the probabilistically weighted average,which provides the adaptive mechanism. The first filter is constant velocity filter (CVF)which is used to estimate the position and velocity of the moving target in non maneuveringcourse. The second filter calculates the acceleration and the new adjustment for the CVF.The second filter is referred as variable velocity filter (VVF). Monte Carlo computersimulation results are included to demonstrate the effectiveness of the proposed algorithm inenhancement the multiple model adaptive filtering.