Ellipsoidal set filter combined set-membership and statistics uncertainties for bearing-only maneuvering target tracking

State estimation of a maneuvering target by bearing-only observations which suffer from set-membership and statistics uncertainties, an ellipsoidal set based filter is proposed in this paper. The influence of two uncertainties is expressed in terms of a set of probability distributions with an ellipsoidal set of means and a covariance matrix. We choose the generalization radius of ellipsoid as the optimality criterion to obtain “tightly” outer bounding ellipsoids. Through simulations of two-state maneuvering tracking problem with bearing-only maneuvering target tracking using two static platforms, proposed estimator has been compared with the extend set-membership filer and the extend Kalman filter. Results illustrate the effectiveness of the ellipsoidal set based filter.