自适应单神经元网络在耦合系统中的应用

基于神经网络PID控制和解耦系统的特点,应用一种针对多输入多输出解耦系统的单神经元PID解耦算法和结构,并在算法中增加了自适应学习速率,更好的实现对耦合系统的解耦控制。仿真结果表明这种单神经元PID解耦算法实现的控制比常规的PID解耦不仅具有较好的动态性能和稳态性能,而且还具有很强的自学习功能和自适应解耦能力。

Application of self- adapting neural network in coupling system

Based on PID neural network control and the characters of decoupling systems, arithmetic and structure of single neural PID decoupling is proposed .The arithmetic can be use in multiple input and multiple output decoupling systems and can realize decoupling control for coupling systems .The simulated results show that this kind of arithmetic is better than the general PID decoupling .It not only has better dynamic performance and stability but also has better self-learning function and self-adapting decoupling ability.