摘要
研究了用神经网络 PID控制器对基于比例阀的气动伺服系统进行控制的方法。用神经网络辨识器来逼近非线性动力学系统 ,并在线修改控制参数。实验及分析表明 ,适当的选择网络参数 ,经过充分的离线训练 ,该控制器可以进行在线的自适应控制 ,系统的控制精度和动态特性有明显提高 ,且在环境参数变化时 ,控制器具有在线自学习和自整定参数的能力。
In this paper a neural network PID controller for proportional valve based pneumatic servo system is designed .A neural network identifier (NNI) is used to approach the nonlinear dynamic system and adjust control parameters on-line. Experiments and analysis indicate that through proper choosing the network parameters and training the network off-line sufficiently the controller can be used to control the plant adaptively on-line, and the control accuracy and dynamic characteristics of the system is increased obviously. Moreover the controller has ability of self learning and self adjusting control parameters when the condition varies.
出处
《中国机械工程》
EI
CAS
CSCD
北大核心
2001年第12期1412-1414,共3页
China Mechanical Engineering
基金
国家自然科学基金资助项目 ( 5 9375 183)