基于C-R模糊模型的广义预测控制算法  

Algorithm of Generalized Predictive Control Based on C-R Fuzzy Model

作　　者：翟春艳[1] 李书臣[1] 

机构地区：[1]辽宁石油化工大学信息学院自动化系,辽宁抚顺113001

出　　处：《计算机仿真》2004年第11期80-81,共2页Computer Simulation

摘　　要：该文对非线性系统的建模采用Cao -Ress(C -R)模糊模型 ,并用卡尔曼滤波算法在线辨识模糊模型的结论参数 ,从而减少了参数辨识的数量和避免了矩阵的求逆运算 ,然后在每一个采样点对该系统进行局部动态线性化 ,根据得到的系统线性化模型对系统采取广义预测控制 (GPC)方法得到当前的控制动作。A Cao-Ress(C-R) fuzzy model is constructed for nonlinear system, and the consequence parameters identification is obtained on-line by using Kalman filtering algorithm, so the number of parameters' identification is reduced and the calculation of matrix inversion is avoided. Local dynamic linearization is applied to the system at each sampling point. Then control action is gained using generalized predictive control(GPC) based on the linearized model. Its effectiveness for nonlinear systems is verified via simulation study.

关 键 词：模糊模型 广义预测控制 非线性系统 卡尔曼滤波算法 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]