Tracking of object using optimal adaptive Kalman filter

Abstract

in this paper we present an optimization based adaptive Kalman filter method is proposed for tracking of an object. In this process noise variance and measurement noise variance are unknown and there is also some error in state of the system. In traditional Kalman filter it is dead beat to find the optimal value of noise variances. In this paper we use innovation based adaptive filtering for estimation of noise variances and memory attenuated filtering is used for state estimation. The proposed adaptive Kalman filter is demonstrated by a example to track an object.