网络控制系统随机鲁棒控制律设计方法  

Stochastic Robust Control Law Design for Networked Control Systems

作　　者：彭琛[1,2] 代冀阳[1] 刘星[2] 吴森堂[2] 

机构地区：[1]南昌航空大学信息工程学院,南昌330063 [2]北京航空航天大学自动化科学与电气工程学院,北京100191

出　　处：《系统仿真学报》2010年第8期1955-1959,2000,共6页Journal of System Simulation

摘　　要：研究了具有随机网络诱导时延及数据包丢失的网络控制系统的随机鲁棒控制律设计方法。利用有限状态markov链来描述控制网络中固有的随机传输延迟和丢包过程,建立网络系统的多结构markov跳变模型。基于由蒙特卡罗估计和优化算法构成的随机鲁棒设计方法,结合保证系统均方稳定的充要条件,得到网络控制系统的随机鲁棒控制律。最后,某型无人飞行器纵向俯仰通道的网络控制仿真实例验证了该方法的有效性。The design method was studied for the stochastic robust control law of the networked control system （NCS） with the characteristics of random-delays and packet-loss.The NCS was modeled as multi-structure Markov jump system,in which the status of random-delays and packet-loss were described as finite-state Markov chains.Considering the sufficient and necessary conditions of the mean-square stability,the controller of the NCS was designed by the stochastic robustness design method based on Monte Carol estimate and optimal algorithm.The numerical simulation of the longitudinal pitch channel for a UAV shows the effectiveness of the proposed approach.

关 键 词：网络控制系统 随机鲁棒设计 马尔科夫跳变线性系统 随机稳定性 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]