COVARIANCE ANALYSIS ALGORITHM FOR INTERCONNECTED SYSTEMS

Abstract

We present in this paper a covariance analysis algorithm for propagation of signal statistics in arbitrarily interconnected nonlinear hybrid systems, applied to six-degree-of-freedom systems specifically in our work. The algorithm uses statistical linearization theory to linearize the nonlinear subsystems, and the resulting linearized subsystems are considered in the original interconnection framework for propagation of the signal statistics. We shall refer to some nonlinearities commonly encountered in six-degree-of-freedom space-vehicle models to illustrate the limitations of this method, along with problems not encountered in standard deterministic simulation analysis. Moreover, we shall exhibit the performance of the algorithm numerically by comparing results from such techniques to results from Monte-Carlo analysis results, both applied to a simple two-dimensional space-intercept problem.