High-gain disturbance observer tuning seen as a multicriteria optimization problem

The paper analyses the potential use of a Non-dominated Sorting Genetic Algorithm (NSGA-II) in finding the parameters of an Extended State Observer (ESO). The ESO in the proposed framework is used as a part of a disturbance-rejection controller, which governs the horizontal position of a nonlinear cart-like system. The considered multicriteria optimization NSGA-II algorithm is introduced to automatically find a set of the observer design parameters that guarantee a desired (and predefined) behavior of the plant. The validity of the considered approach is verified with real experiments conducted on a laboratory testbed.