基于BP神经网络的穿浪双体船姿态自适应PID控制

Adaptive PID Control of Wave Piercing Catamaran Ride Based on BP Neural Network

穿浪双体船能够在一般高速船无法航行的海况中正常执行任务,减小其纵向运动能够更好地提高性能。论文建立了带有T型水翼的穿浪双体船纵向运动控制模型,根据控制模型的特点,针对纵摇和垂荡两个通道设计了PID控制器。基于此,利用BP神经网络,对PID控制参数进行在线整定,建立了参数自适应的PID控制器,并针对两种控制器的控制效果进行了仿真。仿真结果表明,基于BP神经网络的参数自适应能够克服传统PID控制的缺陷,减摇控制效果更好。

Wave piercing catamarans can execute tasks normally in the sea condition where a general high speed vessel cannot sail. Reducing the longitudinal motion can improve its performance better. In this paper,the longitudinal motion control model of wave piercing catamaran with T-foil is established. The PID controller is designed for the two channels of pitching and heaving according to the characteristics of the control model. Based on this,the BP neural network is used to adjust PID control parameters online,the adaptive controller is established and the control effects of the two controller are simulated. The simulation results show that the parameter self-adaption based on BP neural network can overcome the weaknesses of the traditional PID control,whose effect of reducing the vibration is better.