楊冬梅 李達(dá)
摘 要:將耗散理論的二次型供給率中的矩陣Q推廣到正定的情況。進(jìn)而研究了在狀態(tài)轉(zhuǎn)移概率未知的情況下一類(lèi)連續(xù)時(shí)間非線性廣義馬爾可夫跳變系統(tǒng)的嚴(yán)格耗散控制問(wèn)題。在應(yīng)用范圍更廣的Willems耗散性定義的基礎(chǔ)上,首先基于一類(lèi)Lyapunov函數(shù),給出了相應(yīng)的隨機(jī)容許的條件,然后設(shè)計(jì)導(dǎo)數(shù)比例反饋控制器,通過(guò)一系列的矩陣構(gòu)造和合同變換,將雙線性矩陣不等式(BMI)轉(zhuǎn)化為可用LMI工具箱解決的線性矩陣不等式(LMI)。最后通過(guò)數(shù)值算例并結(jié)合Matlab給出實(shí)例,證明其可行性。
關(guān)鍵詞:非線性廣義馬爾可夫跳變系統(tǒng);轉(zhuǎn)移概率部分未知;耗散控制;P-D反饋
Abstract:The matrix Q in the quadratic supply rate of dissipative theory is extended to the case of positive definite. Furthermore, the strictly dissipative control problem for a class of continuous time nonlinear singular Markov jump systems with unknown state transition rates is studied. Based on the more widely used definition of Willems dissipativity, firstly, based on a class of Lyapunov functions, the stochastically admissible conditions are given, and then the proportional derivative feedback controller is designed. Through a series of matrix construction and contract transformation, bilinear matrix inequality (BMI) is transformed into linear matrix inequality (LMI) which can be solved by LMI toolbox. Finally, a numerical example is given to prove its feasibility.
Key words:nonlinear singular Markov jump systems; partly unknown transition rates; strict dissipativity; P-D state feedback
近年來(lái),馬爾可夫系統(tǒng)由于可以更好的描述復(fù)雜系統(tǒng)而受到廣泛關(guān)注[1]。耗散系統(tǒng)理論在廣義系統(tǒng)中也有諸多的應(yīng)用[2],由于耗散性存在的一般性引起學(xué)者關(guān)注。一些學(xué)者考慮了廣義馬爾可夫系統(tǒng)具有耗散性質(zhì)的穩(wěn)定性問(wèn)題[3-5],文獻(xiàn)[6][7]基于Q <0得到一些線性矩陣不等式(LMI)解決了廣義馬爾可夫的輸出耗散反饋控制相關(guān)問(wèn)題。文獻(xiàn)[8][9]討論在狀態(tài)轉(zhuǎn)移概率未知時(shí),考慮了近似求解雙線性不等式(BMI)的方法和量化輸入輸出反饋的相關(guān)問(wèn)題。文獻(xiàn)[10]將耗散控制理論應(yīng)用在一類(lèi)非線性隨機(jī)系統(tǒng)上,得到相關(guān)結(jié)果。但是,在非線性廣義馬爾可夫跳變系統(tǒng)耗散性的研究方面比較少,特別是當(dāng)二次型供給率中的矩陣Q>0的情況,是有待進(jìn)一步研究的。
研究了在狀態(tài)轉(zhuǎn)移概率未知下一類(lèi)非線性廣義馬爾可夫跳變系統(tǒng)的嚴(yán)格耗散問(wèn)題。基于廣義馬爾可夫系統(tǒng)隨機(jī)容許的充分條件,并針對(duì)耗散系統(tǒng)二次型供給率中的矩陣Q>0的情形,給出了將不能用Schur補(bǔ)引理和線性矩陣不等式方法的雙線性矩陣不等式轉(zhuǎn)化為可用LMI工具箱的思路。最后通過(guò)數(shù)值算例,證明結(jié)論有效。
1 問(wèn)題描述與準(zhǔn)備知識(shí)
考慮下面連續(xù)時(shí)間廣義馬爾可夫跳變系統(tǒng):
5 結(jié) 論
研究了非線性廣義馬爾可夫跳變系統(tǒng)的耗散控制問(wèn)題。通過(guò)構(gòu)造一系列的矩陣,然后進(jìn)行合同變換,得到了使系統(tǒng)能夠漸近穩(wěn)定且嚴(yán)格耗散的充分條件,并可用LMI工具箱求解。也給出了控制器設(shè)計(jì)方法。最后通過(guò)數(shù)值算例證明其有效性。
參考文獻(xiàn)
[1] 趙建喆, 吳辰鈮, 王興偉, 等. 基于馬爾可夫毯的貝葉斯網(wǎng)絡(luò)結(jié)構(gòu)學(xué)習(xí)算法[J].東北大學(xué)學(xué)報(bào)(自然科學(xué)版),2020,41(04):464-469.
[2] 黎虹,史書(shū)慧.基于PI控制器的不確定廣義系統(tǒng)的耗散控制[J].控制工程,2018,25(09):1634-1637.
[3] 李景浩.若干類(lèi)廣義馬爾可夫跳變系統(tǒng)的穩(wěn)定性分析[D].沈陽(yáng):東北大學(xué),2013.
[4] LI X, JIANG N. Strictly dissipative control for Markovian jump descriptor systems[C]// Chinese Control Conference. IEEE,2016:1804-1807.
[5] FU L, MA Y C.Dissipative filtering for singular Markov jump systemswi the generally uncertain transition ratesvianew integral inequality approach[J].Journal of the Franklin Institute,2018,355(15):7354-7383.
[6] SEOK I, PARK P. Output feedback control for descriptor fuzzy systems with Markovian jumps[C]// European Control Conference. IEEE,2019:25-28.
[7] PARK C, PARK P.Dynamic output-feed-back dissipative control for singularmarkovian jump systems[C]// Conference of the Society of Instrument and Control Engineers of Japan. IEEE,2018:584-589.
[8] CHEN J, LIN C, CHEN B, et al. Output feedback control for singular Markovian jump systems with uncertain transition rates[J]. IET Control Theory & Applications, 2016,10(16):2142-2147.
[9] 孫維陽(yáng),劉雨.一類(lèi)轉(zhuǎn)移概率部分未知的Markov跳躍系統(tǒng)的輸入輸出量化反饋控制[J].系統(tǒng)工程與電子技術(shù),2019,41(08):1858-1864.
[10]LIN Z, LIU J, NIU Y. Dissipative control of nonlinear stochastic systems with Poisson jumps and Markovian switchings[J]. IET Control Theory & Applications,2012,6(15):2367-2374.