Hybrid estimation algorithms

The optimal, mean-square estimate of the state of a hybrid system is difficult to determine because the equations of state evolution are nonlinear and non-Gaussian. When there is a direct, albeit noisy, measurement of the modal state, it is possible to derive a useful approximation to the optimal estimator. This simplified algorithm is tested on a target tracking problem, and is seen to be superior to the conventional extended Kalman filter.This research was partially supported by a grant from the Hughes Aircraft Company and by the MICRO Program of the State of California under Project No. 91-156.