Kalman Filter aided by Online Estimation on Q and its Application on Target Tracking

Abstract

The Kalman filter is widely used in various fields, however the unknown process noises covariance will decrease the performance of Kalman filter, even lead to filter divergence. To avoid this, an innovative Kalman filter aided by online estimation on unknown process noise covariance is proposed in this paper. It estimates the unknown covariance from the observation sequence by recursive computation. Finally a simulation on target tracking system is given to verify the efficiency and reliability of proposed algorithm.