摘要
系统研究了基于神经网络的离散变结构控制系统设计方法,提出了几种具体设计方案.神经网络的引入可以使滑模(变结构)控制具备学习与自适应能力,使控制信号得以柔化,从而能够减轻或避免困扰常规滑模控制器的抖振现象,改善控制效果.
Discrete variable structure control systems based on neural networks are researched. Some detailed design schemes are given. The introduction of neural networks makes the variable structure control system have learning and adaptive abilities. This means that the control signals can become “soft”, and the chattering phenomenon, which puzzles general sliding mode controllers, can be alleviated, thus the control effects can be improved.
出处
《海军工程大学学报》
CAS
北大核心
2005年第6期48-52,57,共6页
Journal of Naval University of Engineering
关键词
离散时间系统
变结构控制
神经网络
滑模控制
discrete time systems
variable structure control
neural networks
sliding-mode control
