网络系统中动态矩阵控制的稳定性条件  

Stability condition for dynamic matrix control in networked system

作　　者：葛阳[1] 王景成[2] 

机构地区：[1]上海交通大学自动化系,上海200240 [2]系统控制与信息处理教育部重点实验室,上海200240

出　　处：《控制理论与应用》2012年第7期949-954,共6页Control Theory & Applications

基　　金：国家自然科学基金资助项目(60934007,61174059);教育部新世纪人才计划资助项目(NCET-08-0359);上海市青年科技启明星跟踪计划资助项目(11QH1401300)

摘　　要：网络控制系统中的时滞和丢包会影响控制器的稳定性;因此,传统的动态矩阵控制器设计方法无法适应网络化环境.本文针对这一问题设计了一种改进型动态矩阵控制器.通过建立缓存器将不确定时滞转化为固定时滞;并提出了一种更新系统单位阶跃响应系数的方法,用来处理时滞对控制器的影响.采用一次性传输整个控制序列的方法,避免了丢包时需要更新控制量的问题.此外,进一步分析了该控制系统的稳定性问题,给出了考虑时滞和丢包信息的系统稳定性充分条件.最后通过实时仿真软件示范了如何确定允许的最大时滞和丢包.验证了该方法的有效性.The time-delay and the packet-dropout in networked control systems （NCSs） affect the stability of the control system. Hence, the traditional dynamic matrix control （DMC） controller is no longer suitable in the networked environment. To deal with this problem, we designed an improved dynamic matrix controller （DMC）, in which the uncertain time-delay is transformed to the fixed time-delay by two buffers. An updating method for the step-response coefficient is put forward to handle the effect of time-delay on controller. The entire control sequence is transmitted in one time to eliminate the required updating for control data when packet-dropout is occurred. Furthermore, the stability problem of the improved DMC system is investigated and a sufficient condition considering the information of time-delay and packet-dropout is established. By using a real-time simulation software, we illustrate how to determine the maximum tolerable time-delay and packet-dropout for a given example. Results validate the proposed algorithm.

关 键 词：网络控制系统 动态矩阵控制 时滞 丢包 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]