State Estimation of Jump Parameter Systems with State-Dependent Observation Noise

Abstract

A least-squares state estimation algorithm is obtained for a general class of stochastic systems having 1) system nonlinearity ; 2) an unknown jump parameter and 3) state-dependent observation noise. The algorithm developed is consistent in the sense that for each special case with the properties 1) and 2) or 3), it reduces to the algorithm the authors have already developed. Illustrative examples of numerical computation are given for better understanding of the result.