摘要
球磨机制粉系统的出口温度和入口负压控制回路具有比较强的耦合作用,并且对象具有纯延迟特性,PID参数的整定比较复杂。因此,文中提出了基于对角神经网络(DRNN)的解耦控制方案,能够在线自适应的调整PID控制器的3项参数,仿真结果表明可以达到较为理想的控制效果。
In a coal-pulverizing system,the control loop of exit temperature and entry underpressure has a strong coupling effect.Besides,the control objects have the characteristic of large delay time,so the tuning of PID parameters becomes very difficult.In this paper,a decoupling system based on a diagonal recurrent neural network is presented.The system can realize the self-tuning of the three parameters of PID.The result of simulation shows that the system can achieve ideal control.
出处
《工业仪表与自动化装置》
2005年第2期22-25,共4页
Industrial Instrumentation & Automation
关键词
球磨机
PID控制
解耦控制
对角神经网络
ball pulverizer
PID control
decoupling control
diagonal recurrent neural network
