Design of a Kalman filter with Doppler radar measurements

A new algorithm is developed to incorporate Doppler radar measurements into the Kalman filter. An equivalent formulation of the ideal linear measurement equations is derived from the decomposition of measurement noise covariance matrix. Based on the direction cosines estimation and error compensation, a practical measurement equation and sequential filtering equations are given. This sequential process structure can inherit in the direction cosines a performance improvement from updating position measurements and thus reduce its estimation error. Monte Carlo simulation results show that the sequential filtering algorithm can not only improve the estimation accuracy and its performance is superior to conventional extended Kalman filter.