State Estimation of a Non-linear Hybrid System Using an Interacting Multiple Model Algorith

Abstract

In this work, we formulate a state estimation scheme for a nonlinear hybrid system that is subjected to stochastic state disturbances and measurement noise using an interacting Multiple-Model Algorithm (IMM). In particular, we propose the use of an IMM Extended Kalman Filter (IMM-EKF) and an IMM Unscented Kalman filter (IMM-UKF), which belongs to the class of derivative free estimators to carry out estimation of state variables of hybrid system. The efficacy of the proposed state estimation schemes is demonstrated by conducting simulation studies on a two-tank hybrid system. Analysis of the simulation results reveals that the proposed state estimation schemes are able to generate fairly accurate filtered estimate of state variables.