A fast optimization algorithm of observer trajectory for bearings-only target localization

In the mobile sensor network, the relative position of observer and target is an important factor which significantly affects the localization performance of target. In order to improve the localization accuracy of target, optimization algorithm of observer trajectory is proposed. It is based on minimizing the filtered mean square (MS) position error. It estimates the target position utilizing an extended Kalman filter (EKF). It decreases the uncertain region of observer optimization position, based on the observer and target bearing distributing relationships. Simulation results show that the convergence rate, utilizing more observations moving to estimate the target localization, is more quicker than using one observer, and the localization error becomes less. Finally, the most general moving rule of one and more observers is given.