| dc.description.abstract | This thesis presents the application and evaluation of Moving Horizon Estimation (MHE) for the nonlinear two-tank system. MHE is an iterative optimization-based approach that continuously updates the estimates of the states by solving an optimization problem over a fixed-size, receding horizon. Linear and nonlinear MHE-based estimators are designed and implemented in Matlab for evaluation in simulation environment and Simulink for on-line realization and validation. The linear and nonlinear MHE are evaluated in comparison with the Kalman and Extended Kalman filter through extensive simulations and experimental validation, assessing their accuracy, efficiency, and overall performance. The results of the two-tank state and unmeasured disturbance estimation shows the benefit of the MHE. | |
