数值界不确定关联大系统Plug-and-play分散鲁棒模型预测控制  

Plug-and-play Decentralized Robust Model Predictive Control for Large-scale Interconnected Systems with Uncertain Value Bound

作　　者：李艳芬 刘晓华 LI Yanfen;LIU Xiaohua(School of Mathematics and Statistics Science,Ludong University,Yantai 264039,China)

机构地区：[1]鲁东大学数学与统计科学学院,山东烟台264039

出　　处：《鲁东大学学报（自然科学版）》2018年第2期105-111,共7页Journal of Ludong University:Natural Science Edition

基　　金：国家自然科学基金(61174097)

摘　　要：针对一类状态矩阵、控制矩阵存在数值界不确定性的关联大系统,研究了plug-and-play(PNP)分散鲁棒预测控制器设计问题.首先通过有界实引理,将不确定性转化成线性矩阵不等式;然后对于大系统之间的关联项,采用PNP操作,求解优化问题得到分散鲁棒预测控制器;最后数值仿真算例证明了算法的有效性.For a class of large-scale interconnected systems with uncertain value bound in state and control matrices,the design of plug-and-play(PNP)decentralized robust predictive controller was investigated.First of all,the uncertainty was converted into linear matrix inequality(LMI)by bounded real Lemma.Then,for the correlation between large-scale systems,PNP operation was adopted to solve the optimization problem and the decentralized robust predictive controller was obtained.Finally,the numerical simulation shows the validity of the algorithm.

关 键 词：数值界不确定 PNP控制 分散控制 鲁棒模型预测控制 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]