The Application of Self-Tuning Neural Network PID Controller on Longitudinal Motion Reduction in Head Waves

High-speed ships tend to generate large longitudinal motions in waves, leading bow slamming and seasickness. To reduce severe longitudinal motions, actively controlled T-foils are much needed. In this study, a PID controller based on BP neural network is applied to actively control the flap angle of the T-foil. The controller has the self-tuning ability to predict appropriate PID gains. The motion reduction effect of the controller is tested under the condition of regular waves and irregular waves. Compared with the situation where the flap angle of the T-foil is fixed at 0 deg, the maximum pitch can be reduced by 67.8% and 61.9% with the PID gains predicted by the network after several rounds of training in the case of regular and irregular wave incidence, respectively. The results demonstrate that the BP neural network can effectively improve the determination efficiency and robustness of the PID gains.