状态耦合大规模系统的分布式预测控制  

Distributed model predictive control for large scale systems with state-coupled dynamics

作　　者：潘红光[1] 米文毓 高海南[2] 江能前[3] PAN Hong-guang;MI Wen-yu;GAO Hai-nan;JIANG Neng-qian(College of Electrical and Control Engineering,Xi’an University of Science and Technology,Xi’an 710054,China;Xi’an Satellite Control Center,Weinan 714000,China;School of Economics and Finance,Xi’an Jiaotong University,Xi’an 710049,China)

机构地区：[1]西安科技大学电气与控制工程学院,陕西西安710054 [2]西安卫星测控中心,陕西渭南714000 [3]西安交通大学经济金融学院,陕西西安710049

出　　处：《西安科技大学学报》2018年第3期490-497,共8页Journal of Xi’an University of Science and Technology

基　　金：国家自然科学基金(61603295);中国博士后基金(2017M623207);陕西省科技攻关计划(2016GY-040)

摘　　要：研究带有状态耦合的大规模系统的分布式预测控制,这类大规模系统一般可由若干子系统构成。在每个子控制器中均考虑其他子控制器的状态轨迹,这些状态轨迹的假设值可基于历史数据获得。为限制各子系统假设预测轨迹和真实轨迹之间的不确定性偏差加入了相容性约束,并通过在线收紧此约束使得其带来的保守性更小;为消除不确定性偏差,通过对未来轨迹约束的逐步收紧保证物理约束在未来时刻均被满足的;针对真实系统在线优化性能指标,具有更优的控制性能。最后,采用3个弹簧-阻尼小车子系统相互连接而成的系统,通过与已有算法比较,得到了文中算法的计算量更小,控制性能更优的结论,证明了所提方法的有效性。This paper studies the distributed model predictive control（ DMPC） for the large-scale system composed of a set of subsystems with coupled states. In each sub-controller,the state trajectories of the other sub-controllers are involved,which are assumed based-on the solutions at the previous sampling time. A compatibility constraint is imposed to bound the uncertain difference between the assumed predictive trajectory and the actual one,and the conservatism brought by the compatibility constraint is smaller through online tightening this constraint. The constraints on the predicted future evolutions are tightened to counteract this uncertain difference,so that the physical constraint is satisfied at each sampling time. At last,a cart-damper-spring system with three subsystems is adopted to get the conclusion that,the presented algorithm has smaller computing burden and better control performance through comparing with the existing algorithm. The effectiveness of the presented algorithm is certified.

关 键 词：分布式预测控制 动态耦合子系统 约束收紧 相容性约束 

分 类 号：TP13[自动化与计算机技术—控制理论与控制工程]