基于α-置信度网络化控制系统的转换控制  

α-credibility-based switched control for networked control systems

作　　者：马长林[1] 方华京[2] 

机构地区：[1]华中师范大学计算机科学系,湖北武汉430079 [2]华中科技大学控制科学与工程系,湖北武汉430074

出　　处：《华中科技大学学报（自然科学版）》2008年第4期48-50,共3页Journal of Huazhong University of Science and Technology(Natural Science Edition)

基　　金：国家自然科学基金资助项目(60274014,60574088);高等学校博士学科点专项科研基金资助项目(20020487006)

摘　　要：以多步延迟网络化控制系统的控制器设计为基础,提出了一种基于α-置信度的转换控制法.针对时延的两种不同情况定义了平凡和非平凡两种状态:系统所处状态为平凡状态的概率为1-α,该状态下采用使系统均方指数稳定的LQG最优控制器;系统所处状态为非平凡状态的概率为α,该状态下采用使系统随机渐近稳定的时滞补偿控制器.简化了较大延迟情况下LQG最优控制器的求取,工程上易于实现,提高了系统的性能.考虑到两种控制器的综合作用,建立了基于α-置信度的系统转换控制模型,仿真结果证明了转换控制法比单一控制法具有更好的控制效果.On the basis of controller design of networked control systems with multi-step delay, a switched control method based on α-credibility is proposed. Normal and abnormal states are defined for delay＇s two different cases. The probability of the system in normal state is 1-α. LQG optimal controller is adopted in normal state, which is shown to render the system mean square exponentially stable. The probability of the system in abnormal state is a. Time-delay compensation controller is α- dopted in abnormal state, which is shown to render the system stochastic asymptotically stable. The derivation of LQG optimal controller with long delay is simplified and easy to be realized in engineering. The performance of the system is improved. The synthetic roles of two controllers are considered to establish the switched control model of the system based on α-credibility. The result is proved by simulations that the switched control method has better control effects than single control method.

关 键 词：网络化控制系统 多步时延 α-置信度 转换控制 

分 类 号：TP278[自动化与计算机技术—检测技术与自动化装置]