-
自20世纪末开始,网络被逐渐引入控制系统中,由此,网络控制系统应运而生。通常“网络化控制”有两种不同的理解:从计算机网络角度,是对网络自身的控制(control of network),如对网络路由、网络流量等的调度和控制;从自动控制角度,是通过网络对系统的控制(control through network),具体是控制系统的各组成单元(如传感器、执行器、控制器)通过网络完成数据的传输。网络化控制的出现体现了控制系统向智能化、网络化、分布化、集成化的趋势发展。网络控制系统(NCSs)最早于1998年出现在文献[1]中,是通过实时网络将分布在空间不同位置的组成单元链接起来组成的闭环反馈控制系统。与传统的点对点控制系统相比,网络控制系统有成本低、易于维护、系统可靠性高、灵活性高、可实现信息资源共享及远程操作等优点[2]。然而,随着网络控制系统带来极大便利的同时,也给系统的分析与综合带来了巨大的挑战[3-4]。复杂网络控制系统如智能电网[5]、大数据[6]、云控制系统[7]和遥操作系统[8]等复杂网络控制系统的兴起,使网络控制系统具有广域、宽范围、大数据等新特点。由于受网络带宽及承载能力的限制,数据传输不可避免地受时延、时序错乱、数据丢包、网络抖动、时钟同步、误码等问题的影响。这些问题的存在会影响到网络控制系统的性能、分析与设计。其中,网络诱导时延和数据丢包是网络控制系统中最具挑战性的问题。它们不仅会降低系统的性能,还会破坏系统的稳定性。
近年来,为了解决网络控制系统中时延和数据丢包问题,许多控制方法被提出。主要解决方法有:状态增广法[9-10]、模糊逻辑调节法[11-12]、摄动法[13-14]、随机最优控制法[15-16]、鲁棒控制法[17-18]、缓冲队列法[19-20]、预测控制法[21-27]等。其中预测控制是一种非常有效的网络时延和数据丢包补偿算法,它的大致思想是,控制器利用被控对象的模型信息和以往的采样信号估计当前和将来的被控对象的状态信息或输出信息,然后利用估计的信息控制被控对象,从而达到补偿时延和(或)丢包对NCSs的影响。本文梳理了网络控制系统预测补偿算法的研究现状及重要成果,并进一步展望了网络控制系统的发展方向。
-
20世纪70年代,现代控制理论在航空航天等多个高端工业领域取得了辉煌成就的同时,工业控制对象结构、参数和环境的复杂变化,也向控制理论提出了新的挑战。随着对非线性、不确定系统的研究,工业控制过程大量存在非线性、不确定、多输入-多输出的高维复杂系统,传统的控制算法已经不能满足工业生产的需求。实际的工业生产迫切需要一种新的控制技术,人们开始寻找各种对模型要求低、控制综合质量好、在线计算方便的控制算法。预测控制就是在这种背景下发展起来的。
-
如图 4所示,典型的预测控制以对象的阶跃或脉冲响应为模型,采用滚动优化的方式在线地对控制对象实现控制,主要包括[34]:预测模型、滚动优化、反馈校正3个环节。它的基本原理是:首先,根据控制的需求、网络的环境及控制策略等建立预测模型,预测系统未来的输出状态;用受控对象的输出与预测模型输出之间的误差作为反馈信息进行校正,其输出再与参考状态进行比较之后进行滚动优化,然后再计算当前时刻加于系统的控制,并输出给受控单元,完成整个预测控制的循环。其中:
1) 预测模型:即描述系统动态行为的模型,它具有预测功能,能够根据系统现在时刻的控制输入以及过程的历史信息,预测控制对象未来的输出值。如状态方程、传递函数、微分方程、差分方程等都可选作预测模型。
2) 滚动优化:预测控制是一种优化控制算法,它通过某一性能指标的最优化来确定未来的控制作用。如采用跟踪某一期望参考轨迹的二次范数最小,如式(1)所示:
$$\begin{align} & \min J=E\{\sum\limits_{j={{M}_{1}}}^{{{M}_{2}}}{{{[{{y}_{\text{ref}}}(k+j)-y(k+j)]}^{2}}}+ \\ & \sum\limits_{j=1}^{{{M}_{3}}}{{{\delta }_{j}}\Delta u{{(k+j-1)}^{2}}}\} \\ \end{align}$$ (1) 式中,${{y}_{\text{ref}}}(k+j)$、$y(k+j)$分别表示未来$(k+j)$时刻系统的期望和实际输出;M1、M2和M3分别表示最小输出长度、预测长度和控制长度;$\Delta u(k+j-1)$为控制输入增量;${{\delta }_{j}}$表示加权系数。
3) 反馈校正:预测控制是一种闭环控制算法,必须经过反馈校正才能确保滚动优化的有效实施。为了防止模型失配或环境干扰引起控制对理想状态的偏离,在进行预测控制时,新的采样时刻到来时,立即要检测出系统的实际输出值,然后根据实时数据的反馈来校正预测的模型,最后实施新一轮的优化控制。
典型的预测控制算法有3种:动态矩阵控制(dynamic matrix control,DMC)、模型算法控制(model algorithm control,MAC)、广义预测控制(generalized predictive control,GPC)。然而,随着智能控制技术的发展及智能传感器的广泛应用,预测控制开始向智能和数字化方向发展。近几年来,国内外许多学者提出了新的预测控制算法,如数据驱动预测控制、模糊预测控制、神经网络预测控制、基于观测器的预测控制、基于卡尔曼滤波的预测控制等。
-
如图 5所示是具有时延和丢包的网络控制系统结构图。
图 5中,反馈通道和前馈通道时延分别为${{\tau }^{\text{sc}}}(k)$和${{\tau }^{\text{ca}}}(k)$,数据丢包可以用开关s1和s2的断开来表示。网络预测控制算法的基本思想是[35],在传感器端设置缓冲器,将采集的状态数据x(k)或输出数据y(k)和时间戮(time-stamp)一起打包封装,然后发送给控制器。在控制器端设置预测生成器,预测生成器利用接收到的信息预测系统当前和未来时刻的控制输入$\{u(k),u(k+1),\cdots ,u(k+\tau )\}$,然后将预测控制输入序列和时间戮(time-stamp)一起打包发送给执行器。在执行器端设置补偿器,补偿器将接收并存储最新的控制输入序列,同时选择最新的控制输入作为控制对象的输入。
-
广义预测控制(GPC)采用受控自回归积分滑动平均模型(controlled auto-regressive integrated moving-average,CARIMA),该模型本身具有的预测功能,使其可依据系统过去时刻信息和未来输入数据,预估未来时刻的系统输出值[36-38]。广义预测算法通常采用如下差分方程:
$${\boldsymbol A} ({{q}^{-1}})y(k)=B({{q}^{-1}})u(k)+C({{q}^{-1}})e(k)/\Delta $$ (2) 式中,${\boldsymbol A}({{q}^{-1}})$、${\boldsymbol B}({{q}^{-1}})$、${\boldsymbol C}({{q}^{-1}})$为${{q}^{-1}}$的多项式;$u(k)$为系统输入;$y(k)$为系统输出;$e(k)$为白噪声;${{q}^{-1}}$为差分算子。引入丟番图方程:${{\boldsymbol E}_{j}}({{z}^{-1}}){\boldsymbol A}({{z}^{-1}})\Delta +{{z}^{-j}}{{\boldsymbol F}_{j}}({{z}^{-1}})=1$,通过丟番图方程的多步求解运算,预测当前k时刻后系统的输出,并依据预测输出求出相应的控制量。同时,广义预测模型需要先采用相应的辨识方法对未知参数进行正确的辨识。文献[36-37]提出了一种基于广义预测网络时延补偿算法,并将其用在广域电力系统中,设计广义阻尼控制器。文献[38]中,GPC被用于解决网络控制系统中的时延和丢包问题。
-
$${\boldsymbol Y}(k)={{\boldsymbol Y}_{0}}+\bar{\boldsymbol A}\Delta {\boldsymbol U}(k)$$ (3) 式中,${{\boldsymbol Y}_{0}}(k)$和${\boldsymbol Y}(k)$分别表示初始值向量和预测输出向量;$\Delta {\boldsymbol U}(k)$是控制增量序列;$\bar {\boldsymbol A}$表示参数矩阵。基于动态矩阵的时延和丢包预测补偿的基本思想是,对连续状态空间模型$\dot{x}(t)={\boldsymbol A}x(t)+{\boldsymbol B}u(t)$,将其离散化,同时考虑NCS中时延和丢包的影响,可得如下离散数学模型:
$$x(k+1)={\mathit{\pmb{\Phi}}} x(k)+{\mathit {\pmb{\Gamma}}_{0}}({{\tau }_{k}})u(k-l+1)+{\mathit {\pmb{\Gamma}}_{1}}({{\tau }_{k}})u(k-l)$$ (4) 式中,${\mathit{\pmb{\Phi}}} ={{\text{e}}^{Ah}}$;${\mathit {\pmb{\Gamma}}_{0}}({{\tau }_{k}})=\int_{\text{ }0}^{\text{ }lh-{{\tau }_{k}}}{{{\text{e}}^{As}}{\boldsymbol B}\text{ d}s}$;${\mathit {\pmb{\Gamma}}_{1}}({{\tau }_{k}})=\int_{l\text{ }h-{{\tau }_{k}}}^{\text{ }h}{{{\text{e}}^{As}}{\boldsymbol B}\text{ d}s}$;h表示采样周期;l为控制序列长度;${{\tau }_{k}}$表示时延和丢包总和。此时,预测模型的参数矩阵为${\mathit{\pmb{\tilde{A}}}}=f({\mathit{\pmb{\bar{A}}}},{\mathit {\pmb{\Gamma}}_{0}},{\mathit {\pmb{\Gamma}}_{1}})$。文献[39]针对网络控制系统中时延大于一个采样周期及数据丢包的情况,提出了一种改进的动态矩阵控制算法来补偿时延的影响,同时,当数据传输过程中出现丢包时利用动态矩阵算法计算控制量及未来输出预测值的冗余信息代替丢失的信息。文献[40]针对时延大于一个采样周期,但落在某个确定周期范围内的情况,提出一种自适应动态矩阵算法有效补偿网络传输中的延时。文献[41]提出了一种基于动态矩阵的控制算法,有效补偿了传感器到控制器及控制器到执行器的时延和丢包。
-
$$\left\{ \begin{align} & {\boldsymbol x}(k+1)={\boldsymbol A}x(k)+{\boldsymbol B}u(k) \\ & {\boldsymbol y}(k)={\boldsymbol C}x(k) \\ \end{align} \right.$$ (5) 式中,x(k)、u(k)、y(k)分别为系统状态向量、输入向量和输出向量。基于观测器的预测控制算法的主要思想是考虑网络诱导时延和数据丢包可能存在反馈通道和前馈通道,但系统的状态不可观测,于是构造状态观测器对系统的状态进行估计,通过对状态的向前预测来补偿网络传输过程的时延和丢包。文献[42]构造了如下状态观测器:
$${\pmb{{\hat{x}}}_{k+1|k}}=\boldsymbol A{\pmb{{\hat{x}}}_{k|k-1}}+\boldsymbol B{{u}_{k}}+\boldsymbol L({{y}_{k}}-\boldsymbol C{\pmb{{\hat{x}}}_{k|k-1}})$$ (6) 式中,${\pmb{{\hat{x}}}_{k+1|k}}\in {{R}^{n}}$和${{\boldsymbol u}_{k}}\in {{R}^{m}}$分别表示一步状态预测和观测器在k时刻的输入。假设控制器的当前时刻为k,反馈通道网络时延为${{\tau }^{\text{sc}}}$(为了便于理解这里只考虑常时延,同时,数据丢包被看成是一种特殊的时延),基于系统式(5)和观测器式(6)依次预测得:${\pmb{{\hat{x}}}_{k-{{\tau }^{\text{sc}}}+1|k-{{\tau }^{\text{sc}}}}},{\pmb{{\hat{x}}}_{k-{{\tau }^{\text{sc}}}+2|k-{{\tau }^{\text{sc}}}}},\cdots ,{\pmb{{\hat{x}}}_{k|k-{{\tau }^{\text{sc}}}}}$,若同时考虑前馈通道时延${{\tau }^{\text{ca}}}$,同理,预测状态从$k+1$到$k+{{\bar{\tau }}^{\text{ca}}}$时刻状态值(${{\bar{\tau }}^{\text{ca}}}$表示前馈通道时延上界),即${\pmb{{\hat{x}}}_{k+1|k-{{\tau }^{\text{sc}}}}},{\pmb{{\hat{x}}}_{k+2|k-{{\tau }^{\text{sc}}}}},\cdots ,{\pmb{{\hat{x}}}_{k+{{{\bar{\tau }}}^{\text{ca}}}|k-{{\tau }^{\text{sc}}}}}$。基于上述结果,设计控制器:${{\boldsymbol u}_{k+{{{\bar{\tau }}}^{\text{ca}}}|k-{{\tau }^{\text{sc}}}}}={\boldsymbol K}{\pmb{{\hat{x}}}_{k+{{{\bar{\tau }}}^{\text{ca}}}|k-{{\tau }^{\text{sc}}}}}$。其中,K表示控制器增益。文献[43-44]在上述观测器式(6)的基础上,结合网络信道中的时延和丢包提出了新的观测器。文献[44]针对反馈通道存在网络时延问题,在观测器式(3)的基础上提出了一种新形式的观测器,然后通过预测补偿网络控制系统中的时延。
-
数据驱动自适应预测控制的主要思想是,针对一个未知的离散时间非线性系统,利用动态线性化数据模型方法,在闭环系统的每个动态工作点建立一个虚拟等价的动态线性化数据模型,然后基于该虚拟等价的数据模型设计控制器。该方法仅依赖于被控系统实时量测的数据,利用受控系统的数据进行控制器设计和分析,不依赖受控系统任何的数学模型信息,不需要任何外在的测试信号[45-48]。文献[45]提出一种基于数据的预测控制算法,有效补偿了双通道的数据丢包。文献[46]提出了一种无模型自适应控制算法(MFCA)对网络控制系统中的数据丢包进行预测补偿。文献[47]针对网络控制系统的时延问题,提出了一种数据驱动预测控制算法有效补偿了时延对网络控制系统的影响。文献[48]提出一种基于数据的预测控制算法,用于网络控制系统的时延和丢包补偿。
-
模糊模型可用于表示复杂的非线性系统,将模糊模型与预测控制相结合形成模糊预测控制。许多学者将模糊预测控制理论应用于网络控制系统中,并提出了基于模糊预测控制的时延和丢包补偿算法。模糊预测控制算法的基本思想是,首先,通过一组模糊规则将非线性控制系统表示成一组模糊线性系统,然后用预测控制算法对模糊系统中的时延和丢包进行补偿[49-50]。文献[49-50]中,非线性控制系统被表示为Takagi-Sugeno(T-S)模糊模型。文献[49]设计了基于模糊模型的状态和输出反馈控制器,文献[50]设计了基于模糊模型的状态和积分控制器,然后用该控制器对反馈通道和前馈通道中的时延进行预测补偿。
-
神经网络(NN)对复杂的非线性函数具有非常强的逼近能力,它可以通过对历史数据的训练预测出系统未来时刻的动态。将神经网络与预测控制算法相结合形成预测控制器,利用已知的数据预测出系统当前的状态,可以实现对网络控制系统中的时延和丢包进行补偿[51-53]。在文献[51]中,BP神经网络作为预测控制器被用来补偿网络控制系统中的时延。文献[52]将自适应神经网络和预测控制相结合组成自适应预测控制器,对网络控制系统中的时延和丢包进行补偿。
-
基于卡尔曼滤波(Kalman filtering)估计的预测控制算法的基本思想是,把卡尔曼滤波估计器作为网络控制系统中的预测生成器对系统中的状态进行估计,从而达到对时延和丢包的补偿功能[54-56]。文献[54]用卡尔曼滤波估计算法补偿网络控制系统中的时延。文献[55]用卡尔曼滤波预测补偿算法对反馈通道的时延和数据丢包进行了有效的补偿。
-
基于Smith预估器补偿算法的基本思想是,针对闭环特征方程含的影响系统控制品质的滞后问题,在反馈通道引入Smith估计器,消除闭环特征方程中时延的影响,以提高系统的稳定性[57-59]。文献[57-58]将Smith预估器分别与神经网络和Fuzzy-PI控制相结合组成新的估计器补偿网络中的时延。文献[59]提出一种自适应Smith估计器有效克服了网络中时延的影响。
-
文献[60-62]中,模型预测控制(MPC)算法被用于补偿网络传输中的时延和丢包,MPC属于典型预测控制算法的一种。它的基本思想是,通过建立一个描述系统动态的模型且该模型具有预测功能,用其预测系统的未来输出,同时补偿NCS中时延和丢包。文献[60]设计了包含改进型模型预测控制(MPC)和Smith预估器的新型预测控制方法,由MPC计算得到的预测控制量补偿前向通道时延,Smith预估器的作用是补偿反馈通道时延,同时还设计了一个滤波器提高网络控制系统的鲁棒性。
Overview and Research Trends of Predictive Control Method for Network Control Systems
-
摘要: 随着智能电网、大数据、云控制系统和遥操作系统等复杂网络控制系统的兴起,网络控制系统具有了广域、宽范围、大数据等新特点。因此,解决网络诱导的时延和数据丢包问题在复杂网络控制系统中更显重要。该文对网络诱导的时延和数据丢包进行了分析,对网络预测控制算法研究现状进行了分析和总结,并展望了网络控制系统未来的发展方向。该文旨在为复杂网络控制系统的预测控制提供理论基础和研究思路。Abstract: With the emergence of complex network control systems (NCSs), such as smart grid, big data, cloud control system and teleoperation system, NCSs possess the new characteristics of wide area, wide range, and large data etc. This paper focuses on the network induced-delay and packet dropouts in the complex NCSs, which need more attention to. The research status of predictive control algorithm is examined and summarized, and also the future development directions of network control systems are discussed. This paper aimed to provide theoretical basis and research ways to predictive control of the complex network control systems.
-
Key words:
- delay /
- network control systems (NSCs) /
- networked predictive control /
- packet dropouts
-
[1] WALSH G C, YE H, BUSHNELL L G. Stability analysis of networked control systems[J]. IEEE Transactions on Control Systems Technology, 2002, 10(3):438-446. doi: 10.1109/87.998034 [2] LIU Kun, EMILIA F, KARL H J. Dynamic quantization of uncertain linear networked control systems[J]. Automatica, 2015(59):248-255. https://www.researchgate.net/publication/282275552_Dynamic_quantization_of_uncertain_linear_networked_control_systems [3] QIU Li, YAO Feng-qi, XU Gang, et al. Output feedback guaranteed cost control for networked control systems with random packet dropouts and time delays in forward and feedback Communication Links[J]. IEEE Transactions on Automation Science and Engineering, 2016, 13(1):284-295. doi: 10.1109/TASE.2014.2353657 [4] QIU Jian-bin, GAO Hui-jun, STEVEN X D. Recent advances on fuzzy-model-based nonlinear networked control Systems:a survey[J]. IEEE Transactions on Industrial Electronics, 2016, 63(2):1207-1217. doi: 10.1109/TIE.2015.2504351 [5] WANG Shao-bu, MENG Xiang-yu, CHEN Tong-wen. Wide-area control of power systems through delayed network communication[J]. IEEE Transactions on Control Systems Technology, 2012, 20(2):495-503. doi: 10.1109/TCST.2011.2116022 [6] EMIL J K, RAQUEL B, MUNOZ P, et al. Self-healing in mobile networks with big data[J]. IEEE Communications Magazine, 2016, 54(1):114-120. doi: 10.1109/MCOM.2016.7378435 [7] XIA Yuan-qing. Cloud control systems[J]. IEEE/CAA Journal of Automatica Sinica, 2015, 2(2):134-142. doi: 10.1109/JAS.2015.7081652 [8] CHO M H, LEE C H. A low-power real-time operating system for ARC (actual remote control) wearable device[J]. IEEE Transactions on Consumer Electronics, 2010, 56(3):1602-1609. doi: 10.1109/TCE.2010.5606303 [9] ZHU Xiao-yuan, ZHANG Hui, WANG Jun-min, et al. Robust lateral motion control of electric ground vehicles with random network-induced delays[J]. IEEE Transactions on Vehicular Technology, 2015, 64(11):4985-4995. doi: 10.1109/TVT.2014.2383402 [10] HALEVI Y, RAY A. Integrated communication and control system:Part-analysis[J]. Journal of Dynamic Systems, Measurement and Control, 1988, 110(4):367-373. doi: 10.1115/1.3152698 [11] JIANG X F, HAN Q L. On designing fuzzy controllers for a class of nonlinear networked control systems[J]. IEEE Transactions on Fuzzy Systems, 2008, 16(4):1050-1060. doi: 10.1109/TFUZZ.2008.917293 [12] ZHANG H G, YANG J, SU C Y. T-S fuzzy-model-based robust H∞ design for networked control systems with uncertainties[J]. IEEE Transactions on Industrial Informatics, 2007, 3(4):289-301. doi: 10.1109/TII.2007.911895 [13] WALSH G C, BELDIMAN O, BUSHNELL L G. A symptotic behavior of nonlinear networked control systems[J]. IEEE Transactions on Automatic Control, 2001, 46(7):1093-1097. doi: 10.1109/9.935062 [14] ZHANG Chao-long, DENG Fei-qi, PENG Yun-jian, et al. Adaptive synchronization of Cohen-Gross berg neural network with mixed time-varying delays and stochastic perturbation[J]. Applied Mathematics and Computation, 2015, 269(15):792-801. [15] WU J, CHEN T W. Design of networked control systems with packet dropouts[J]. IEEE Transactions on Automatic Control, 2007, 52(7):1314-1319. doi: 10.1109/TAC.2007.900839 [16] HUANG D, GUANG S K. State feedback control of uncertain networked control systems with random time delays[J]. IEEE Transactions on Automatic Control, 2008, 53(3):829-834. doi: 10.1109/TAC.2008.919571 [17] GKTAS F. Distributed control of systems over communication networks[D]. Philadelphia, Pennsylvania, USA:University of Pennsylvania, 2000. [18] NĔSIĆ D, TEEL A R. Input-to-state stability of networked control systems[J]. Automatica, 2004, 40(12):2121-2128. http://www.researchgate.net/publication/222755176_Input-to-state_stability_of_networked_control_systems [19] LUCK R, RAY A. An observer-based compensator for distributed delays[J]. Automatica, 1990, 26(5):903-908. doi: 10.1016/0005-1098(90)90007-5 [20] LUCK R, RAY A. Experimental verification of a delay compensation algorithm for integrated communication and control systems[J]. International Journal of Control, 1994, 59(6):1357-1372. doi: 10.1080/00207179408923135 [21] WANG Rui, WANG Bo, LIU Guo-ping, et al. H∞ Controller design for networked predictive control systems based on the average dwell-time approach[J]. IEEE Transactions on Circuits and Systems-Ⅱ:Express Briefs, 2010, 57(4):310-314. doi: 10.1109/TCSII.2010.2043386 [22] LIU Yuan-ming, FONG I-kong. Robust predictive tracking control of networked control systems with time-varying delays and data dropouts[J]. IET Control Theory Application, 2013, 7(5):738-748. doi: 10.1049/iet-cta.2012.0603 [23] ZOU Y, CHEN T, LI S. Network-based predictive control of multirate systems[J]. IET Control Theory Appl, 2010, 4(7):1145-1156. doi: 10.1049/iet-cta.2008.0577 [24] ZHANG Jin-hui, XIA Yuan-qing, SHI Peng. Design and stability analysis of networked predictive control systems[J]. IEEE Transactions on Control Systems Technology, 2013, 21(4):1495-1501. doi: 10.1109/TCST.2012.2208967 [25] PANG Zhong-hua, LIU Guo-ping. Design and implementation of secure networked predictive control systems under deception attacks[J]. IEEE Transactions on Control Systems Technology, 2012, 20(5):1334-1342. doi: 10.1109/TCST.2011.2160543 [26] TANG P L, SILVA C W. Compensation for transmission delays in an ether net-based control network using variable-horizon predictive control[J]. IEEE Transactions on Control Systems Technology, 2006, 14(4):707-718. doi: 10.1109/TCST.2006.876640 [27] WANG Shao-ying, FANG Hua-jing, TIAN Xue-gang. Recursive estimation for nonlinear stochastic systems with multi-step transmission delays, multiple packet dropouts and correlated noises[J]. Signal Processing, 2015(115):164-175. https://www.researchgate.net/publication/276853049_Recursive_estimation_for_nonlinear_stochastic_systems_with_multi-step_transmission_delays_multiple_packet_dropouts_and_correlated_noises [28] MOHAMED E M B G, ARBEN Ç, YSKANDAR H. Optimal integrated control and scheduling of networked control systems with communication constraints:application to a car suspension system[J]. IEEE Transactions on Control Systems Technology, 2006, 14(4):776-787. doi: 10.1109/TCST.2006.872504 [29] ABDELKADER A, POLUSHIN I G., ABDELHAMID T. Synchronization of nonlinear systems with communication delays and intermittent information exchange[J]. Automatica, 2015(59):1-8. http://dl.acm.org/citation.cfm?id=2825876 [30] ZHU Qi-xin, LIU Hong-li, HU Shou-song. Uniformed model of networked control systems with long time delay[J]. Journal of System Engineering and Electronics, 2008, 19(2):385-390. doi: 10.1016/S1004-4132(08)60097-3 [31] WANG Yue, XIA Meng, GUPTA V, et al. On feedback passivity of discrete-time nonlinear networked control systems with packet drops[J]. IEEE Transactions on Automatic Control, 2015, 60(9):2434-2439. doi: 10.1109/TAC.2014.2384754 [32] RABELLO A, BHAYA A. Stability of asynchronous dynamical systems with rate constraints and applications[J]. IEEE Proceeding Control Theory Application, 2003, 150(5):546-550. doi: 10.1049/ip-cta:20030704 [33] SCHENATO L. To zero or hold control inputs with lossy links[J]. IEEE Transactions on Automatic Control, 2009, 54(5):1093-1099. doi: 10.1109/TAC.2008.2010999 [34] HUANG Y L,HELEN H L, GONG J P, et al. Fuzzy model predictive control[J]. IEEE Transactions on Fuzzy Systems, 2000, 8(6):665-667. doi: 10.1109/91.890326 [35] LIU G P, XIA Y Q, REES D, et al. Design and stability criteria of networked predictive control systems with random network delay in the feedback channel[J]. IEEE Transactions on Systems, Man, and Cybernetics:Part C, 2007, 37(2):173-184. doi: 10.1109/TSMCC.2006.886987 [36] YAO Wei, JIANG Lin, WEN Jin-yu, et al. Networked predictive control based wide-area supplementary damping controller of SVC with communication delays compensation[C]//2013 IEEE, Power and Energy Society General Meeting (PES). Vancouver, BC:IEEE:1-5. [37] YAO Wei, JIANG Lin, WEN Jin-yu, et al. Wide-area damping controller for power system interarea oscillations:a networked predictive control approach[J]. IEEE Transactions on Control Systems Technology, 2015, 23(1):27-36. doi: 10.1109/TCST.2014.2311852 [38] HU Wen-shan, LIU Guo-ping, DAVID R. Event-driven networked predictive control[J]. IEEE Transactions on Industrial Electronics, 2007, 54(3):1603-1613. doi: 10.1109/TIE.2007.894720 [39] 葛阳, 王景成. 基于动态矩阵的随机时延丢包网络控制系统[J]. 信息与控制, 2011, 40(6):835-840. http://www.cnki.com.cn/Article/CJFDTOTAL-XXYK201106020.htm GE Yang, WANG Jing-cheng. Networked control system with random delay and packet loss based on dynamic matrix[J]. Information and Control, 2011, 40(6):835-840. http://www.cnki.com.cn/Article/CJFDTOTAL-XXYK201106020.htm [40] 费春国, 李昌刚, 张奇智, 等. 网络控制系统中自适应DMC算法的仿真研究[J]. 系统仿真学报, 2008, 20(1):65-68. http://www.cnki.com.cn/Article/CJFDTOTAL-XTFZ200801018.htm FEI Chun-guo, LI Chang-gang, ZHANG Qi-zhi, et al. Simulation of adaptive DMC algorithm in networked control systems[J]. Journal of System Simulation, 2008, 20(1):65-68. http://www.cnki.com.cn/Article/CJFDTOTAL-XTFZ200801018.htm [41] YANG Li-man, LIU Gui-lin, GUO Zhong-wei. Dynamic matrix control algorithm for networked control systems with delay and data packet dropout[C]//2010 IEEE Conference on Cybernetics and Intelligent Systems. Singapore:IEEE, 2010:179-184. [42] LIU Guo-ping, XIA Yuan-qing, CHEN Jie, et al. Networked predictive control of systems with random network delays in both forward and feedback channels[J]. IEEE Transactions on Industrial Electronics, 2007, 54(3):1282-1297. doi: 10.1109/TIE.2007.893073 [43] HA Q P, NGUYEN D T, PHAN T N, et al. Partial state estimation for linear systems with output and input time delays[J]. ISA Transactions, 2014, 53(2):327-334. doi: 10.1016/j.isatra.2013.12.025 [44] ZHANG Jin-hui, LIN Yu-juan, SHI Peng. Output tracking control of networked control systems via delay compensation controllers[J]. Automatica, 2015, (57):85-92. https://www.researchgate.net/publication/276852955_Output_tracking_control_of_networked_control_systems_via_delay_compensation_controllers [45] PANG Zhong-hua, LIU Guo-ping, ZHOU Dong-hua, et al. Data-based predictive control for networked non-linear systems with two-channel packet dropouts[J]. IET Control Theory Appl, 2015, 9(7):1154-1161. doi: 10.1049/iet-cta.2014.0745 [46] HOU Zhong-sheng, BU Xu-hui. Model free adaptive control with data dropouts[J]. Expert Systems with Applications, 2011, 38(8):10709-10717. doi: 10.1016/j.eswa.2011.01.158 [47] XIA Yuan-qing, XIE Wen, LIU Bo, et al. Data-driven predictive control for networked control systems[J]. Information Sciences, 2013, 235(20):45-54. http://www.researchgate.net/publication/256720694_Data-driven_predictive_control_for_networked_control_systems [48] PANG Zhong-hua, LIU Guo-ping, ZHOU Dong-hua, et al. Data-based predictive control for networked nonlinear systems with network-induced delay and packet dropout[J]. IEEE Transactions on Industrial Electronics, 2016, 63(2):1249-1257. doi: 10.1109/TIE.2015.2497206 [49] ZHANG Jin-hui, SHI Peng, XIA Yuan-qing. Fuzzy delay compensation control for T-S fuzzy systems over network[J]. IEEE Transactions on Cybernetics, 2013, 43(1):259-268. doi: 10.1109/TSMCB.2012.2204744 [50] ZHANG Hui, SHI Yang, LIU Ming-xi. Step tracking control for networked discrete-time nonlinear systems with integral and predictive actions[J]. IEEE Transactions on Industrial Informatics, 2013, 9(1):337-345. doi: 10.1109/TII.2012.2225434 [51] YI Jian-qiang, WANG Qian, ZHAO Dong-bin. BP neural network prediction-based variable period sampling approach for networked control systems[J]. Applied Mathematics and Computation, 2007, 185:976-988. doi: 10.1016/j.amc.2006.07.020 [52] WANG Tong, GAO Hui-jun, QIU Jian-bin. A combined adaptive neural network and nonlinear model predictive control for multirate networked industrial process control[J]. IEEE Transactions on Neural Networks and Learning Systems, 2016, 27(2):416-425. doi: 10.1109/TNNLS.2015.2411671 [53] ZHANG Dan, YU Li, WANG Qing-guo, et al. Estimator design for discrete-time switched neural networks with asynchronous switching and time-varying delay[J]. IEEE Transactions on Neural Networks on Learning Systems, 2012, 23(5):827-834. doi: 10.1109/TNNLS.2012.2186824 [54] SHI Ling, XIE Li-hua, RICHARD M M. Kalman filtering over a packet-delaying network:a probabilistic approach[J]. Automatica, 2009(45):2134-2140. http://www.docin.com/p-1388093550.html [55] PROBST A M E, MAGANA O, SAWODNY O, et al. Using a Kalman filter and a Pade' approximation to estimate random timedelays in a networked feedback control system[J]. IET Control Theory Application, 2010, 4(11):2263-2272. doi: 10.1049/iet-cta.2009.0339 [56] HAN Chun-yan, ZHANG Huan-shui, FU Min-yue. Optimal filtering for networked systems with Markovian commun-ication delays[J]. Automatica, 2013(49):3097-3104. http://cn.bing.com/academic/profile?id=2126338352&encoded=0&v=paper_preview&mkt=zh-cn [57] ZHANG W, BRANICKY M S, PHILLIPS S M. Stability of networked control systems[J]. IEEE Control Systems Magazine, 2001, 21(1):84-99. doi: 10.1109/37.898794 [58] CHEN C H, LIN C L, HWANG T S. Stability of networked control systems with time-varying delays[J]. IEEE Communications Letters, 2007, 11(3):270-272. doi: 10.1109/LCOMM.2007.060716 [59] LAI Chen-liang,HSU Pau-lo. Design the remote control system with the time-delay estimator and the adaptive Smith predictor[J]. IEEE Transactions on Industrial Informatics, 2010, 6(1):73-80. doi: 10.1109/TII.2009.2037917 [60] MU Jun-xia, LIU G P, DAVID R. Design of robust networked predictive control systems[C]//Proceedings of American Control Conference. Portland, OR, USA:IEEE, 2005:638-643. [61] GIUSEPPE F, FRANCESCO T, DOMENICO F. Model predictive control for constrained networked systems subject to data losses[J]. Automatica, 2015(54):272-278. [62] PANG Zhong-hua, LIU Guo-ping, ZHOU Dong-hua, et al. Output tracking control for networked systems:a model-based prediction approach[J]. IEEE Transactions on Industrial Electronics, 2014, 61(9):4867-4877. doi: 10.1109/TIE.2013.2289890