Application of multiple model estimation to a recursive terrain height correlation system

This paper describes the results of an investigation of the performance capabilities of an extended Kalman filter (EKF)-based recursive terrain correlation system proposed for low-altitude helicopter navigation. The major disadvantage of this concept is its sensitivity to initial position error. One method for reducing this sensitivity, involves the use of multiple model estimation techniques. In the multiple model approach, a bank of identical EKF's, each of which is initialized at a different point in the a priori uncertainty basket, is employed to ensure that one filter is initialized near the true aircraft position. In this manner, the probability of filter convergence is increased substantially, leading to improved navigation performance.