Adaptive Two-Step Filter with Applications to Bearings-Only Measurement Problem
Di Zhou, Chundi Mu, Wenli Xu · Journal of Guidance Control and Dynamics · 1999
Introduction T HE state estimation of a nonlinear system is still an important research subject. In some cases, the usual extended Kalman lter (EKF) produces large estimation errors and even diverges. In this paper, a two-step lter1;2 for a class of nonlinear systems that consists of a linear dynamic model and a nonlinear measurement model is introduced. In many practical systems, the statistical properties of measurement noise are time varying and unknown a priori. For this case, a statistical estimator can be used to determine the mean and covariance of measurement noise online. A modi ed SageHusa time-varyingmeasurementnoise statisticalestimator is integratedwith the two-step lter to produce an adaptive two-step lter (ATSF). Finally, the ATSF is applied to the bearingsonly measurement problem. Numerical comparisons of the ATSF with the EKF and the adaptive EKF (AEKF) are made using this application.