Generalized Predictive Control for pi-Type Tudder Anti-pitching Based on Nonlinear Adaptive Intelligent Model

This article proposed a method to build an adaptive intelligent model of ship longitudinal motion aiming at maritime moving in all-weather conditions and ship non-linear longitudinal hydrodynamic coefficients. By taking 442 T surface ship as the research object, it calculated typical sea state hydrodynamic coefficients according to water tank experiment results and strip theory, optimized radial basis function neural network (RBF) by optimal stopping rule, build an adaptive three-dimensional nonlinear intelligence model based on real-time change of speed, sea situations and course angle, and obtained ship longitudinal motion real-time online control model in all-weather conditions, by taking white noise as system disturbance, deduced ship-type rudder anti-pitching generalized predictive control (GPC) law. Simulation results show that t he control effect of anti-pitching l can increased by 34% and GPC control have strong robustness by using adaptive intelligent model combined with GPC.