An improvement on the iterated Kalman filter
Niu Xin-liang, Zhao Guo-qing, Yuanhua Liu, Chang Hong · IET Conference Publications · 2009
An improved iterated Kalman filter (IKF) is proposed to reduce the sensitivity of the filter to the initial estimate error. According to the essence of the IKF, i.e., the Gauss-Newton method is used to approximate a maximum likelihood estimate, a new update method is obtained. Simulations show that the improved IKF has better performance than the IKF when the initial estimate error is large. (4 pages)