Moving Horizon Estimation of Constrained Nonlinear Systems by Carleman Approximations

In this paper we investigate the use of moving horizon state estimation techniques for nonlinear systems in the presence of hard constraints. To be able to apply standard QP-based moving horizon estimation schemes, the Carleman technique is used in order to obtain a linear approximation of the nonlinear system. The performance of the resulting estimation scheme is evaluated by simulation on a well studied example from the literature, and is compared with the performance of two available estimation techniques for nonlinear systems, namely the extended Kalman filter and the polynomial extended Kalman filter. The simulations presented show that the proposed constrained moving horizon estimation scheme compares favourably with the other two techniques when the system states are subject to hard constraints