基于驻留概率信息的网络化切换系统的控制和滤波
发布时间:2018-10-05 21:19
【摘要】:切换系统是由有限的子系统和一个切换规则构成,可以用于许多单一模型无法刻画的复杂系统,较一般模型更加贴近真实系统。另外,随着工程中对控制性能、安全等要求越来越高,通过网络完成系统的监测、控制已经得到广泛研究。因此,网络化切换系统的研究具有重要的研究价值。驻留概率表示在足够长的时间范围内切换系统停留在某个特定子系统的概率。本文提出一种基于驻留概率信息的方法,运用此方法得到新的切换系统模型,通过Lyapunov泛函的方法分析切换系统的输出反馈控制器设计问题及有限时间H∞控制问题,并且在切换系统的基础上引入网络诱导延迟、数据丢包,建立新型的网络化切换系统,分析解决异步切换的网络化切换系统的H∞控制问题和网络信道衰落的系统滤波问题。首先,通过引入驻留概率信息提出一种新的随机切换系统模型,利用多Lyapunov泛函方法和线性矩阵不等式技术给出使系统鲁棒H∞均方稳定的充分条件,并且得到H∞性能抑制水平γ和输出反馈控制器的设计方法,通过给出的仿真实例验证方法的有效性。其次,针对离散切换系统的有限时间H∞控制问题,利用驻留概率信息,给出一种新的切换系统模型,利用Lyapunov泛函方法得到系统随机有限时间H∞有界的充分条件,并给出有限时间H∞反馈控制器的设计。同时,考虑到部分驻留概率未知和系统参数存在不确定性的情况,并给出了相应的分析方法,用一组切换系统的数据针对不同的情况证明方法是正确的。接着,研究了异步网络化切换系统的H∞控制问题。考虑网络诱导延迟和数据丢包,通过具有时变延迟的异步采样切换系统和驻留概率信息得到一种新的网络化切换模型。然后给出使网络化切换系统鲁棒H∞随机稳定的充分条件、H∞性能抑制水平γ和相应的反馈控制器的设计方法,并通过仿真验证所提出的方法可以使系统满足期望的性能。最后,考虑在实际系统中网络的信道衰落问题,通过驻留概率信息的方法建立新的网络化切换系统模型,利用Lyapunov泛函方法,分析了网络化切换系统的H∞滤波性能,并给出H∞滤波器设计方法,运用仿真证明方法的有效性。
[Abstract]:The switched system is composed of a limited subsystem and a switching rule. It can be used in many complex systems which can not be described by a single model, and is closer to the real system than the general model. In addition, with more and more requirements of control performance and safety in engineering, control has been widely studied through network monitoring. Therefore, the research of networked switched systems has important research value. The resident probability represents the probability that the switching system stays in a particular subsystem for a long enough time. In this paper, a method based on dwell probability information is proposed. A new switching system model is obtained by using this method. The output feedback controller design problem and finite time H 鈭,
本文编号:2254899
[Abstract]:The switched system is composed of a limited subsystem and a switching rule. It can be used in many complex systems which can not be described by a single model, and is closer to the real system than the general model. In addition, with more and more requirements of control performance and safety in engineering, control has been widely studied through network monitoring. Therefore, the research of networked switched systems has important research value. The resident probability represents the probability that the switching system stays in a particular subsystem for a long enough time. In this paper, a method based on dwell probability information is proposed. A new switching system model is obtained by using this method. The output feedback controller design problem and finite time H 鈭,
本文编号:2254899
本文链接:https://www.wllwen.com/kejilunwen/zidonghuakongzhilunwen/2254899.html