模糊Hammerstein模型预测控制及其在pH过程中的应用  

Predictive control algorithm based on fuzzy Hammerstein model and its application to pH process

作　　者：马新迎 施华峰[1] 苏成利[1] 李平[1] 

机构地区：[1]辽宁石油化工大学信息与控制工程学院,辽宁抚顺113001

出　　处：《计算机与应用化学》2013年第10期1122-1126,共5页Computers and Applied Chemistry

基　　金：国家自然科学基金项目(61203021);辽宁省科技攻关项目(2011216011)

摘　　要：针对一类Hammerstein模型描述的非线性系统,提出了一种改进的非线性预测控制算法。该算法将T-S模型与动态矩阵控制算法相结合,利用动态阶跃响应描述Hammerstein模型的动态线性部分,以克服参数化模型因阶次辨识不准带来的未建模动态问题。利用T-S模型逼近其非线性部分,通过T-S模型的线性化将控制输入转化为动态模型的输入,进而采用线性预测控制方法求解预测控制律,避免了非线性优化在线求解。pH中和过程仿真结果表明该算法具有良好的跟踪性、抗干扰性和较强的鲁棒性。An improved nonlinear predictive control algorithm is proposed for a class of nonlinear systems which are described by Hammerstein model. The algorithm combines T-S model with dynamic matrix control algorithm and describes the dynamic linear part of the Hammerstein model by the dynamic step response, in order to overcome the un-model dynamic problems which is brought by inaccurate identification of the parametric model. A T-S model is used to approximate the nonlinear section of its model, and the input of control is converted into the input of dynamic model by the linearization of the T-S model, then the predictive control law is solved by using the linear predictive control method. Consequently the problem in solving the nonlinear online optimization is avoided. Simulation results of pH process show that the proposed algorithm has a good tracking performance ,anti-interference and strong robustness.

关 键 词：T-S模型 动态矩阵控制 HAMMERSTEIN模型 PH过程 

分 类 号：TP13[自动化与计算机技术—控制理论与控制工程]