无限时间长时延网络控制系统的随机最优控制  

Infinite time stochastic optimal control of networked control systems with long delay

作　　者：朱其新[1] 胡寿松[1] 刘亚[1] 

机构地区：[1]南京航空航天大学自动化学院

出　　处：《控制理论与应用》2004年第3期321-326,共6页Control Theory & Applications

基　　金：国家自然科学基金项目(60234010);南通工学院自然科学基金项目(院自200303).

摘　　要：考虑二次性能指标下线性网络控制系统的随机最优控制问题,建立了控制器为事件驱动时长时延线性网络控制系统的数学模型,证明了在无限时间情况下离散随机黎卡提代数方程解的存在性,设计出无限时间情况下线性网络控制系统的随机最优控制器,得到相应的最优性能指标的表达形式,并证明了相应的随机最优控制器可使网络控制系统均方指数稳定.最后以网络控制下的倒立摆为对象进行仿真研究,仿真结果表明该方法的正确性和有效性.The stochastic optimal control problem was considered in case of linear networked control systems and quadratic criteria.The mathematic model of linear networked control systems with long delay is set up when the controller is event driven.It is proved that the solution of infinite horizon discrete time stochastic algebraic Riccati equation is existent.The infinite horizon stochastic optimal controller of linear networked control systems is designed.The expression of optimal performance index of networked control systems is given.The paper proves that the infinite horizon stochastic optimal controller of networked control systems can render the networked control systems exponentially mean square stable.Finally,the simulation based on the networked controlled inverted pendulum is studied.The simulation results show the proposed methods are feasible and effecive.

关 键 词：网络控制系统 事件驱动 最优控制 均方指数稳定 

分 类 号：TP13[自动化与计算机技术—控制理论与控制工程]