Brief Constrained linear state estimation-a moving horizon approach
Automatica (Journal of IFAC)
Hi-index | 0.00 |
A solution is proposed to estimate the states in the nonlinear discrete time system. Moving Horizon Estimation (MHE) is used to obtain the approximated states by minimizing a criterion that is the Euclidean form of the difference between the estimated outputs and the measured ones over a finite time horizon. The differential evolution (DE) algorithm is incorporated into the implementation of MHE in order to solve the optimization problem which is presented as a nonlinear programming problem due to the constraints. The effectiveness of the approach is illustrated in simulated systems that have appeared in the moving horizon estimation literature.