Genetic algorithms for multiobjective predictive control

Abstract

Control of nonlinear uncertain dynamical systems is considered. The artificial neural networks (ANNs) are used to model the process. For each operating level an ANN is determined. The model predictive type of controller is designed that utilizes a set of ANN model and employs the input constraints. The nondominated sorting genetic algorithm (NSGA) is applied to solve the multiobjective optimization problem. The proposed control schema is applied to a numerical example and the simulation results are included.