孫家玉 嚴(yán)懷成 李郅辰 詹習(xí)生
摘要 針對(duì)具有馬爾可夫切換信道的兩自由度(2-DOF) 四分之一汽車懸架系統(tǒng),研究了事件觸發(fā) ?H ?∞ 濾波問題.首先,信道切換由馬爾可夫鏈控制;其次,考慮到事件觸發(fā)的通信方案,由于有限的網(wǎng)絡(luò)帶寬,產(chǎn)生信號(hào)量化和隨機(jī)丟包問題;然后,采用馬爾可夫線性跳變系統(tǒng)模型來表示整個(gè)濾波網(wǎng)絡(luò)系統(tǒng).利用Lyapunov泛函和線性矩陣不等式方法將事件觸發(fā) ?H ?∞ 濾波問題轉(zhuǎn)化為凸優(yōu)化問題,從而設(shè)計(jì)了切換信道相關(guān)的濾波器,使得濾波誤差系統(tǒng)在均方意義上是指數(shù)穩(wěn)定的并達(dá)到期望的性能水平.最后,通過仿真實(shí)例驗(yàn)證了所提出的設(shè)計(jì)方法的有效性.
關(guān)鍵詞
事件觸發(fā)傳輸機(jī)制;信號(hào)量化;隨機(jī)丟包;馬爾可夫鏈
中圖分類號(hào)? TP273
文獻(xiàn)標(biāo)志碼? A
0 引言
作為道路交通的主要元素之一,汽車的性能直接影響交通系統(tǒng)中車輛和人員的安全,而作為車輛重要部件的懸架系統(tǒng)的控制和估算策略受到越來越多的關(guān)注.20世紀(jì)30年代初就有人研究了旋轉(zhuǎn)和懸架運(yùn)動(dòng)如何影響乘坐性能的問題.隨著懸架系統(tǒng)的快速發(fā)展,數(shù)學(xué)模擬和計(jì)算機(jī)技術(shù)開始應(yīng)用于懸架系統(tǒng)研究領(lǐng)域.近年來,大量關(guān)于懸架控制系統(tǒng)的研究已經(jīng)出現(xiàn)[1-4] .一方面,由于懸架系統(tǒng)的部件之間的磨損和碰撞,將難以操縱車輛;另一方面,系統(tǒng)的所有狀態(tài)在實(shí)踐中都很難獲得.基于可用的輸出測(cè)量,可以設(shè)計(jì)濾波器來估計(jì)不可測(cè)量的系統(tǒng)狀態(tài).而且,獲得懸架系統(tǒng)的更準(zhǔn)確的狀態(tài)估計(jì)值有助于最大化車輛用戶在車輛安全性、舒適性和操縱穩(wěn)定性方面的滿意度.最近,許多研究人員在有限的時(shí)間范圍內(nèi)集中研究懸架系統(tǒng)的濾波問題[5-7] .其中,文獻(xiàn)[5]提出了一種新的方法來解決在各種路況下使用卡爾曼濾波器(KF)進(jìn)行懸架系統(tǒng)狀態(tài)估計(jì)的問題.仿真結(jié)果表明,所提出的自適應(yīng)KF算法可以獲得懸架系統(tǒng)的高精度狀態(tài)估計(jì).
隨著Internet技術(shù)的快速發(fā)展,被控系統(tǒng)與網(wǎng)絡(luò)通信系統(tǒng)的集成已成為網(wǎng)絡(luò)控制技術(shù)的熱點(diǎn),并提出了網(wǎng)絡(luò)控制系統(tǒng)(NCSs).在傳統(tǒng)控制系統(tǒng)的基礎(chǔ)上,隨著網(wǎng)絡(luò)的加入,系統(tǒng)的性能得到了極大的提高,也不可避免地帶來了以下缺陷:網(wǎng)絡(luò)引起的延遲、丟包以及單包和多包傳輸問題.這些問題的存在不僅會(huì)降低系統(tǒng)的控制性能,而且會(huì)導(dǎo)致系統(tǒng)不穩(wěn)定.實(shí)際上,基于網(wǎng)絡(luò)的系統(tǒng)主要通過無線通信網(wǎng)絡(luò)傳輸數(shù)據(jù),因此,信號(hào)必須在傳輸之前被量化,并且許多文獻(xiàn)已經(jīng)報(bào)道了受到量化的網(wǎng)絡(luò)系統(tǒng)的問題[8-9] .
同時(shí),由于未知的外部干擾,通信環(huán)境易受攻擊,從而導(dǎo)致通信環(huán)境的變化,例如時(shí)變采樣周期、時(shí)變延遲和多個(gè)通信信道(MCC)等問題.考慮的一種實(shí)際情況是網(wǎng)絡(luò)中存在MCC,并假設(shè)信道切換由連續(xù)時(shí)間馬爾可夫過程控制[10-12] .在多個(gè)切換通信信道的背景下,采用馬爾可夫跳變系統(tǒng)模型來表示整個(gè)網(wǎng)絡(luò)系統(tǒng),并且已經(jīng)研究了主動(dòng)懸架系統(tǒng)的濾波問題.由于切換是隨機(jī)的,因此,信道切換是由馬爾可夫鏈控制的.
另外,時(shí)間觸發(fā)的采樣機(jī)制易于執(zhí)行和分析,但從資源利用的角度來看,它不太可取.近年來,文獻(xiàn)[13-16]提出了各種事件觸發(fā)機(jī)制作為最小化通信資源的替代方法,僅在調(diào)用所謂的預(yù)先設(shè)計(jì)的觸發(fā) 條件時(shí)才發(fā)送采樣數(shù)據(jù).因此,借助于事件觸發(fā)機(jī)制,在保持期望性能的同時(shí)提高了通信效率.例如,文獻(xiàn)[16]構(gòu)建了一個(gè)事件觸發(fā)的實(shí)時(shí)調(diào)度程序,它決定在任何給定時(shí)刻應(yīng)該執(zhí)行哪種控制模式.
目前,對(duì)于具有馬爾可夫切換的兩自由度四分之一汽車懸架系統(tǒng)的事件觸發(fā)濾波的研究比較少.因此,本文設(shè)計(jì)了濾波器以實(shí)時(shí)監(jiān)控運(yùn)行在切換信道網(wǎng)絡(luò)環(huán)境中的兩自由度四分之一汽車懸架系統(tǒng)狀態(tài)的問題.本文的主要貢獻(xiàn)有:
1)本文考慮了具有馬爾可夫鏈控制的MCC通信環(huán)境,并提供了一種更通用的通信框架,即事件觸發(fā)機(jī)制,信號(hào)量化和隨機(jī)丟包問題同時(shí)被考慮,這更符合實(shí)際的網(wǎng)絡(luò)化控制系統(tǒng);
2)提出了周期性事件觸發(fā)機(jī)制,它基于相對(duì)誤差的閾值條件和量化方案,保證期望的系統(tǒng)性能水平的同時(shí)提高資源利用率;
3)利用Lyapunov泛函方法,設(shè)計(jì)了與切換信道相關(guān)的濾波器,使得濾波誤差系統(tǒng)在均方意義上是指數(shù)穩(wěn)定的,并達(dá)到規(guī)定的系統(tǒng)性能水平.
1 問題描述
3個(gè)通信信道在上述馬爾可夫鏈下的切換時(shí)序以及系統(tǒng)濾波誤差的動(dòng)態(tài)過程如圖1所示.從圖1可以看出濾波誤差在大約幾步內(nèi)趨向于零,系統(tǒng)對(duì)于外部噪聲有很好的魯棒性.
圖2是狀態(tài) x 1 ,x 2 ,x 3 ,x 4 ?的動(dòng)態(tài)軌跡和對(duì)應(yīng)的濾波值,從仿真結(jié)果可以看出濾波器能夠很好地估計(jì)系統(tǒng)的狀態(tài)值.
事件觸發(fā)的時(shí)序如圖3所示,和一般的周期性采樣方式相比,僅有70%數(shù)據(jù)包被傳輸.因此,通過事件觸發(fā)傳輸機(jī)制,在保證系統(tǒng)性能的同時(shí)節(jié)省了通信資源.
4 結(jié)束語
本文研究了汽車懸架系統(tǒng)在多通道切換的通信環(huán)境下的事件觸發(fā) H ?∞濾波問題,同時(shí)考慮了兩方面的網(wǎng)絡(luò)化通信缺陷:信號(hào)量化和隨機(jī)丟包問題.然后,通過構(gòu)建李雅普諾夫函數(shù)方法導(dǎo)出了濾波器參數(shù)以及事件觸發(fā)參數(shù)的設(shè)計(jì)結(jié)果.最后,對(duì)于懸架系統(tǒng)的仿真研究的結(jié)果證實(shí)了本文提出的理論方法的有效性.未來,將考慮把通道時(shí)延、對(duì)象建模的非線性及不確定性考慮在一個(gè)統(tǒng)一的框架中,對(duì)該框架的濾波問題進(jìn)行研究.
參考文獻(xiàn)
References
[ 1 ]?Li ?H Y,Jing X J,Karimi H R.Output-feedback-based? ?H ?∞ control for vehicle suspension systems with control delay[J].IEEE Transactions on Industrial Electronics,2014,61(1):436-446
[ 2 ] Wang ?R R,Jing H,Karimi H R,et al.Robust fault tolerant? ?H ?∞ control of active suspension systems with finite-frequency constraint[J].Mechanical Systems & Signal Processing,2015,62/63(4702):341-355
[ 3 ] Yan H C,Qian F F,Zhang H,et al. H ?∞ fault detection for networked mechanical spring-mass systems with incomplete information[J].IEEE Transactions on Industrial Electronics,2016,63(9):5622-5631
[ 4 ] Zapateiro ?M,Pozo F,Karimi H R,et al.Semiactive control methodologies for suspension control with magnetorheological dampers[J].ASME Transactions on Mechatronics,2012,17(2):370-380
[ 5 ] Wang ?Z F,Dong M M,Qin Y C,et al.Suspension system state estimation using adaptive Kalman filtering based on road classification[J].Vehicle System Dynamics,2017,55(3):371-398
[ 6 ] Pletschen N,Diepold K J.Nonlinear state estimation for suspension control applications:a Takagi-Sugeno Kalman filtering approach[J].Control Engineering Practice,2017,61:292-306
[ 7 ] Sun ?W C,Pan H H,Gao H J.Filter-based adaptive vibration control for active vehicle suspensions with electrohydraulic actuators[J].IEEE Transactions on Vehicular Technology,2016,65(6):4619-4626
[ 8 ] Zhang D,Xu Z H,Karimi H R,et al.Distributed filtering for switched linear systems with sensor networks in presence of packet dropouts and quantization[J].IEEE Transactions on Circuits & Systems Ⅰ:Regular Papers,2017,64(10):2783-2796
[ 9 ] Fu M Y,Xie L H.The sector bound approach to quantized feedback control[J].IEEE Transactions on Automatic Control,2005,50(11):1698-1711
[10] Yin X Y,Zhang L X,Ning Z P,et al.State estimation via Markov switching-channel network and application to suspension systems[J].IET Control Theory & Applications,2017,11(3):411-419
[11] Yan ?H C,Zhang H,Yang F W,et al.Event-triggered asynchronous ?guaranteed cost control for Markov jump discrete-time neural networks with distributed delay and channel fading[J].IEEE Transactions on Neural Networks and Learning Systems,2018,29(8),3588-3598
[12] Li ?Q,Shen B,Liu Y R,et al.Event-triggered? ?H ?∞ state estimation for discrete-time stochastic genetic regulatory networks with Markovian jumping parameters and time-varying delays[J].Neurocomputing,2016,174:912-920
[13] Zhang ?H,Zheng X Y,Yan H C,et al.Codesign of event-triggered and distributed? ?H ?∞ filtering for active semi-vehicle suspension systems[J].ASME Transactions on Mechatronics,2017,22(2):1047-1058
[14] Yan ?H C,Zhang H,Yang F W,et al.Event-triggered asynchronous ?guaranteed cost control for Markov jump discrete-time neural networks with distributed delay and channel fading[J].IEEE Transactions on Neural Networks and Learning Systems,2018,29(8):3588-3598
[15] Peng C,Han Q L,Dong Y.To transmit or not to transmit:a discrete event-triggered communication scheme for networked Takagi-Sugeno fuzzy systems[J].IEEE Transactions on Fuzzy Systems,2013,21(1):164-170
[16] Zhang H,Wang Z P,Yan H C,et al.Adaptive event-triggered transmission scheme and? ?H ?∞ filtering co-design over a filtering network with switching topology[J].IEEE Transactions on Cybernetics,2018,DOI:10.1109/TCYB.2018.2862828
[17] Zhang D,Wang Q G,Srinivasan D,et al.Asynchronous state estimation for discrete-time switched complex networks with communication constraints[J].IEEE Transactions on Neural Networks & Learning Systems,2018,29(5):1732-1746
Event-triggered ?H ?∞ filtering of car suspension systems with Markovian switching
SUN Jiayu 1 YAN Huaicheng1,2 ?LI Zhichen 2 ZHAN Xisheng 2
1 College of Information Science and Engineering,East China University of Science and Technology,Shanghai 200237
2 College of Mechatronics and Control Engineering,Hubei Normal University,Huangshi 435002
Abstract? An event-triggered? ?H ?∞ state estimation problem is investigated in this paper for a two-degrees-of-freedom(2-DOF) quarter-car suspension system operated over a switching-channel network environment.First,the channelswitching is governed by a Markov chain.Then,a Markov jump linear system model is adopted to represent the overall networked system in accordance with the event-triggered communication scheme,signal quantization,and random packet losses on account of the limited network bandwidth.Using the Lyapunov functional and linear matrix inequality method,the event-triggered? ?H ?∞ state estimation problem is transformed into an optimization problem,theswitching-channel-dependent filters of which are designed such that the filter error system is exponentially stable in the mean square sense and achieves the desired performance level.Finally,a simulation example is used to demonstrate the validity of the proposed design.
Key words? event-triggered mechanism;signal quantization;random packet dropouts;Markov chain
南京信息工程大學(xué)學(xué)報(bào)2018年6期