基于TSA的多变量模糊预测控制  

Multivariable Fuzzy Predictive Control Based on TSA

作　　者：吴佳 施惠元[1] 苏成利[1] WU Jia;SHI Hui-yuan;SU Cheng-li(School of Information and Control Engineering,Liaoning Petrochemical University,Fushun 113001,China;Ansteel Mining Co.,Ltd.Qidashan Iron.Ore,Anshan 114043,China)

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 [2]鞍钢集团矿业有限公司齐大山铁矿,辽宁鞍山114043

出　　处：《控制工程》2021年第12期2456-2461,共6页Control Engineering of China

基　　金：国家自然科学基金资助项目(61803191,61703191);辽宁省联合自然科学基金资助项目(2019-KF-03-05);工业控制技术国家重点实验室开放课题(ICT2021B34);高等学校创新人才支持计划项目(LR2019037)。

摘　　要：针对多输入多输出的非线性系统,提出了一种基于树和种子算法(tree and seed algorithm,TSA)的多变量模糊预测控制算法。该算法首先将非线性系统的过程模型建模为Takagi-Sugeno(T-S)模型,并将其作为预测模型。在此基础上,利用该模型和反馈校正方法,递推得到校正的输出预测值。然后,给定相应的二次性能指标及输入输出约束条件,利用TSA在线搜索多变量非线性预测控制系统的最优控制律,避免了直接递推控制律时复杂多项式难以求解的问题。最后,以中和反应过程为例进行仿真对比实验。实验结果验证了所提算法的有效性和可行性。For the multi-input mutli-output nonlinear systems,a multivariable fuzzy predictive control algorithm based on tree and seed algorithm(TSA)is proposed.Firstly,the process model of the nonlinear system is modeled as a Takagi-Sugeno(T-S)model and used as a predictive model.On this basis,this model and the feedback correction method are used to obtain the corrected output predicted value by recursion.Then,the corresponding quadratic performance index and input and output constraints are given,and the TSA is used to search the optimal control law of multivariable nonlinear predictive control system on-line,which can avoid the problem that the complex polynomials are difficult to solve in the direct recursive control law.Finally,the neutralization reaction process is taken as an example for simulation and comparison experiments.The experimental results verify the effectiveness and feasibility of the proposed algorithm.

关 键 词：非线性 树和种子算法 预测控制 TAKAGI-SUGENO模型 中和反应 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]