论文部分内容阅读
本文针对船舶操纵这种非线性、时变参数控制对象，提出了一种采用神经网络自适应PID控制方案。该控制结构有两个子神经网络组成，一个三层BP神经网络用于对被控对象进行在线辨识，另一个两层线性网络构成具有PID结构的控制器。文中给出了神经网络的在线训练学习方法，并进行了船舶操纵控制仿真研究。
In this paper, a nonlinear neural network adaptive PID control scheme is proposed for ship maneuvering this nonlinear and time-varying parameter control object. The control structure has two sub-neural networks, a three-layer BP neural network for on-line identification of the controlled object and another two-layer linear network to form a controller with PID structure. In this paper, the online training and learning method of neural network is given, and the simulation research of ship steering control is carried out.