含有概率约束的网络化控制系统的区域控制及滤波

在网络化控制系统（Networked Control Systems, NCSs）中，传感、控制以及被控体之间通过网络进行互联,其“传感-控制-驱动-被控体”间的因果关系变得异常复杂。特别地，当NCSs中的变量存在概率约束时，如何揭示并应对随机网络环境在不同时空尺度上对NCSs控制性能的影响，解决其分析与控制问题，是一个亟待解决的科学问题，也是NCS研究的热点与难点问题。本项目针对含有概率约束的网络化控制系统，提出区域控制及区域滤波的科学性问题，即，本项目旨在设计控制器/滤波器，使系统的输出/误差在指定区域停留的概率足够大。由此可以带来更大的控制自由度以及满足更多控制性能指标。在此基础上，本项目将深入探讨概率约束、网络随机特性、控制区域、随机控制器以及停留概率之间的内在科学机理，建立充分反映他们耦合规律的网络化控制系统模型，设计合适的控制器/滤波器来满足区域控制/滤波的目的。

In networked control systems (NCSs), sensors, controllers and plant are connected through network, the causal relationships among “sensor-control-actuator-plant” become much complex. In particular, when there are probabilistic constraints in the physical variables of the NCSs, how to reveal the influence of the random network environments for the control performance, analyze and control the NCSs are hot but difficult points in NCSs research. Considering the probabilistic constraints in NCSs, this project introduces the concept of“region control/filter”, that is, design the controller or filter such that the probability of system output/error falling in a specific region is large enough. Furthermore, this project will also explore the inherent scientific mechanism among the probabilistic constraints, random network properties, control region, the stochastic controller and the sojourn probabilities. Finally, the purpose of this project is to establish a general model based on above issues and design proper controller/filter to guarantee the probabilistic purpose of the region control.