Skip to Main Content
The paper presents a novel adaptive neural-network based nonlinear model predictive control (NMPC) methodology for hybrid systems with mixed inputs. For this purpose an online self-organizing growing and pruning redial basis function (GAP-RBF) neural network is employed to identify the hybrid system using the unscented Kalman filter (UKF) learning algorithm. A receding horizon adaptive NMPC is then devised based on the identified GAP-RBF neural network model. The resulting nonlinear optimization problem is solved by a genetic algorithm (GA). The performance of the proposed adaptive model predictive control methodology is illustrated on a motorboat simulation case study.