基于RBF网络的PID参数自整定控制器在恒压供水系统中的应用  

Parameter self-tuning PID Controller Based on RBF Neural Network for Constant Pressure Water Supply with Frequency Conversion Syste

作　　者：肖佐无[1] 罗麦丰[1] 

机构地区：[1]株洲职业技术学院,湖南株洲412001

出　　处：《工业控制计算机》2012年第11期54-55,共2页Industrial Control Computer

基　　金：湖南省教育厅资助科研项目(10C0345)

摘　　要：针对变频恒压供水对象的非线性、不确定、大滞后的特点,设计了基于RBF网络的PID参数自整定控制器,该控制器能根据RBF网络辨识的Jacobian信息,采用梯度下降算法自动在线调整PID控制器的参数,达到自适应控制的目的,克服了传统PID控制设计中的参数调整困难的问题,同时控制系统的动态和静态性能有较大的提高。通过仿真验证了控制系统的有效性。The constant pressure water supplying system with frequency conversion have characteristics about non-linear,uncer- tain,large delay and so on.The Parameter Self-tuning PID Controller was designd based RBF neural network,Jacobian ma- trix is identified by RBF neural network,the gradient descend method is used to adjust the Parameter of PID controller,the adaptability of the system to the time-variation of the parameters is improved,the disadvantage of difficult parameter ad- justment of the conventional PID control system is got over,performance of control system is improved.

关 键 词：变频恒压供水 RBF网络 PID参数自整定 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]