基于改进灰色预测模型的自整定PID控制  

Self-adjusting PID control based on modified grey predictive model

作　　者：林丽君[1] 王宏楠[1] 李书臣[1] 

机构地区：[1]辽宁石油化工大学控制理论与控制工程学院,辽宁抚顺113001

出　　处：《西安工程科技学院学报》2007年第6期817-820,共4页Journal of Xi an University of Engineering Science and Technology

基　　金：辽宁省教育厅科学研究计划资助项目(2004D031)

摘　　要：针对非线性滞后系统提出一种基于改进灰色预测模型的自整定PID控制.采用变步长四阶龙格-库塔法与灰色模型相结合的方法作为预测模型,通过BP神经网络建立参数自学习PID控制器,根据目标函数E调整Δu,使系统输出最优化.仿真结果证明,该预测控制具有较好的跟踪效果,且有效地克服了误差及干扰等不确定因素所带来的影响.Self-adjusting PID control based on modified grey predictive model is proposed for nonlinear systems with time delay. The step-varying fourth order Runge-Kutta and GM（2,1） grey model are introduced for predictive model. PID controller of self-learning is erected by BP neural network, △u is adjusted by objective function E,so that output of the system is optimum. Simulation result shows that the traceability of predictive control is good and influences coming from error and uncertain factor are retrained.

关 键 词：非线性滞后系统 四阶龙格-库塔法 灰色模型 BP神经网络 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]