Guidance of Seekerless Interceptors Using Innovation Covariance Based Tuning of Kalman Filters

Abstract

In this paper, an integrated estimation/guidance algorithm that combines Kalman-filter-based interactive multiple-model estimators with a modified differential game guidance law is developed for seekerless interceptors in three-dimensional space. The target is assumed to perform various types of maneuvers, while the sensor has noisy measurements and the interceptor is subject to acceleration bound. To handle these, the Kalman filters used in the interactive multiple model are pretuned using two recently developed metrics based on innovation covariance. To determine the filter tuning parameters, the metrics are evaluated offline for the Kalman filter using different process models, and the tuned values of the process noise covariance are selected in each case. These tuned Kalman filters are then used in the interactive multiple-model configuration to cater to various maneuvers that are expected during the end game. A numerical study compares the performances of these, and a previously reported interactive multiple-model technique demonstrates significant improvements in performance.