摘要
提出一种用于船舶航向保持的人工神经网络控制器，设计了一种“直接控制法”，与目前常见的几种人工神经网络控制方法相比，其特点是结构简洁，可直接应用在实时控制中。在实际应用中，只要知道被控过程的定性控制特点，“直接控制法”即可用于干扰环境下的非线性过程控制。“直接控制法”在船舶自动舵上进行了验证，并同传统的ＰＩＤ自动舵进行了比较，结果表明，在随机风和测量噪声的干扰下，人工神经网络自动舵的控制效果明显优于ＰＩＤ自动舵。
A neural network approach for controlling the course of ships is presented.A direct control method which has characteristics of simpler structure and practical applicability for realtime controlling compared with several other common neural network control methods is proposed.In practical application,the direct control method can be used for controlling the nonlinear behavior of the plant under the conditions of disturbances and noise with minor qualitative knowledge about the plant.The direct control method has been tested on the auto navigators of ships.The comparison of performance with a conventional PID has been made.Under random wind forces and measurement noise,results show that the feasibility and adaptive property of the proposed auto navigator with artificial neural networks control methods is absolutely better than the PID controller.
出处
《华东船舶工业学院学报》
1998年第6期36-37,共2页
Journal of East China Shipbuilding Institute(Natural Science Edition)
关键词
人工智能
神经网络
船舶
自动操舵仪
artificial intelligence
neural networks
ships
auto navigators
