Generalized minimum variance unbiased joint input-state estimation and its unscented scheme for dynamic systems with direct feedthrough

This paper proposes a joint input-state estimation technique that can estimate unknown inputs, states and parameters of dynamic systems with direct feedthrough. The estimation is performed in a minimum variance unbiased manner, based on measurements from the dynamic system. This paper also presents a novel unscented minimum variance unbiased (UMVU) estimation scheme to realize the proposed estimator via unscented transform. Five UMVU variants are implemented and applied in the numerical study. The results have shown successful estimation of unknown inputs and states, and accurate identification of unknown system parameters. The characteristics and performance of the five UMVU variants are also demonstrated through the comparison of the results.