基于神经网络的船舶自动舵设计

Automatic ship rudder′s design based on neural network

依据人工操舵的基本原理 ,提出了一种能适应船舶航行环境变化和自动获取船舶操纵运动特性知识的神经网络控制器。仿真结果表明 ,此神经网络控制器不仅能对系统进行有效的控制 ,而且可达到无超调、无波动的控制效果。此外 。

According to the principle of manual steering, the authors suggest a neural network controller which can adapt the change of navigation environment and get the information of ship manuverbility automatically. The simulation results show that the neural network controller not only can control the system effectually ,but also can achieve no overshoot and fluctuation. In addition , the operational procedure of the controller is in a good agreement with the optimal procedure of manual steering.