姜寶續(xù) 卜祥偉 齊強(qiáng)
摘 要: 針對傳統(tǒng)神經(jīng)網(wǎng)絡(luò)預(yù)設(shè)性能控制方法結(jié)構(gòu)復(fù)雜、控制實(shí)時(shí)性不高等問題,為乘波體飛行器(Waverider Vehicle,WV)提出了一種低復(fù)雜度的預(yù)設(shè)性能反演控制方法。首先,設(shè)計(jì)了一種形式簡單、不依賴初始誤差的新型性能函數(shù),對跟蹤誤差進(jìn)行包絡(luò)約束。其次,采用轉(zhuǎn)換誤差策略,通過引入一種轉(zhuǎn)換誤差,將對跟蹤誤差的不等式約束等價(jià)轉(zhuǎn)換為等式約束。然后,基于轉(zhuǎn)換誤差,為WV設(shè)計(jì)了一種無需任何學(xué)習(xí)參數(shù)的新型反演控制器,并避免了傳統(tǒng)反演控制的“微分項(xiàng)膨脹”問題。最后,仿真結(jié)果表明,所提方法可以保證跟蹤誤差良好的動(dòng)態(tài)性能與穩(wěn)態(tài)精度。
關(guān)鍵詞: ?乘波體飛行器;預(yù)設(shè)性能;新型性能函數(shù);新型反演控制器;轉(zhuǎn)換誤差
中圖分類號: TJ765;V448? 文獻(xiàn)標(biāo)識碼: A? 文章編號: ?1673-5048(2021)02-0011-10
0 引? 言
乘波體飛行器(Waverider Vehicle,WV)具有響應(yīng)快速、機(jī)動(dòng)性高、航程遠(yuǎn)、突防能力強(qiáng)、探測與攔截難度大、殺傷力大等突出特點(diǎn),且具備傳統(tǒng)航天飛行器所不具有的戰(zhàn)略、戰(zhàn)術(shù)與效費(fèi)比等方面的突出優(yōu)勢,已經(jīng)成為世界各航空航天大國為爭奪空天權(quán)而爭先發(fā)展的方向[1-4]。
作為WV能夠安全、高效地執(zhí)行和完成飛行任務(wù)的核心,控制系統(tǒng)設(shè)計(jì)目前屬于國際前沿?zé)狳c(diǎn),國內(nèi)外均有眾多學(xué)者和科研機(jī)構(gòu)開展這方面的研究。WV構(gòu)型特殊復(fù)雜,飛行環(huán)境未知因素多,導(dǎo)致為其建立的運(yùn)動(dòng)模型存在很大的非線性和不確定性。同時(shí),WV的超機(jī)動(dòng)飛行對其控制系統(tǒng)的動(dòng)態(tài)性能也提出了極高的要求。因此,控制系統(tǒng)的魯棒性與動(dòng)態(tài)性能也成為了關(guān)注熱點(diǎn)。文獻(xiàn)[5]基于反演控制設(shè)計(jì)結(jié)構(gòu),引入徑向基函數(shù)神經(jīng)網(wǎng)絡(luò)對模型未知函數(shù)和不確定項(xiàng)進(jìn)行逼近,實(shí)現(xiàn)了參數(shù)不確定情形下對WV參考指令的魯棒跟蹤,但沒有考慮神經(jīng)網(wǎng)絡(luò)逼近誤差的影響。進(jìn)一步,文獻(xiàn)[6]利用全局調(diào)節(jié)動(dòng)態(tài)神經(jīng)網(wǎng)絡(luò)進(jìn)行在線逼近WV的未知模態(tài),并引入魯棒項(xiàng)處理神經(jīng)網(wǎng)絡(luò)估計(jì)誤差帶來的影響。文獻(xiàn)[5-6]雖能實(shí)現(xiàn)對WV參考指令的魯棒跟蹤控制,但是對控制系統(tǒng)的動(dòng)態(tài)性能的約束效果不是很理想。為此,文獻(xiàn)[7]基于預(yù)設(shè)性能控制(Prescribed Performance Control,PPC)方法,通過設(shè)計(jì)性能函數(shù)對跟蹤誤差進(jìn)行包絡(luò)約束,可以定量調(diào)節(jié)跟蹤誤差的收斂速度、超調(diào)量等動(dòng)態(tài)性能指標(biāo),從而實(shí)現(xiàn)控制系統(tǒng)滿意的動(dòng)態(tài)性能指標(biāo)。文獻(xiàn)[8]提出一種新型魯棒預(yù)設(shè)性能控制器設(shè)計(jì)方法,能同時(shí)滿足高超聲速飛行器瞬態(tài)響應(yīng)和穩(wěn)態(tài)性能,通過設(shè)計(jì)一種新型時(shí)變對數(shù)型障礙Lyapunov函數(shù),結(jié)合動(dòng)態(tài)面法,保證了WV高度和速度子系統(tǒng)穩(wěn)態(tài)跟蹤誤差精度和動(dòng)態(tài)性能。
以上方法雖能較好地保證控制系統(tǒng)的魯棒性與動(dòng)態(tài)性能指標(biāo),但是采用神經(jīng)網(wǎng)絡(luò)逼近需要對其權(quán)值矩陣的元素進(jìn)行在線學(xué)習(xí),在線學(xué)習(xí)量大,控制實(shí)時(shí)性不高。現(xiàn)有的PPC性能函數(shù)形式過于復(fù)雜,設(shè)計(jì)參數(shù)較多,工程實(shí)用性不高?;诖?,本文為WV設(shè)計(jì)了一種低復(fù)雜度預(yù)設(shè)性能反演控制方法。設(shè)計(jì)一種構(gòu)型簡單的新型性能函數(shù)對跟蹤誤差進(jìn)行約束,基于反演設(shè)計(jì)框架,采用轉(zhuǎn)換誤差設(shè)計(jì)控制器,避免了對虛擬控制律的求導(dǎo)計(jì)算。與傳統(tǒng)神經(jīng)網(wǎng)絡(luò)控制方法相比,所提方法不需要任何在線學(xué)習(xí)參數(shù),可以滿足控制算法良好的實(shí)時(shí)性。最后,通過仿真實(shí)驗(yàn)驗(yàn)證了控制策略的有效性和優(yōu)越性。
1 問題描述
1.1 WV縱向運(yùn)動(dòng)模型
以NASA研制的X-43驗(yàn)證機(jī)為對象,建立WV縱向運(yùn)動(dòng)的二維模型[9],如圖1所示。
建模采用二維激波-膨脹波理論計(jì)算定常氣動(dòng)力與力矩并忽略了粘性影響;在充分考慮剛體運(yùn)動(dòng)與彈性振動(dòng)的影響后采用一階活塞理論近似計(jì)算非定常氣動(dòng)力;采用準(zhǔn)一維定常流動(dòng)假設(shè)建立發(fā)動(dòng)機(jī)模型;采用拉格朗日方程建立了WV的彈性動(dòng)力學(xué)方程;忽略模型中弱耦合與慢動(dòng)態(tài)因素,建立一個(gè)面向控制器的WV縱向運(yùn)動(dòng)參數(shù)擬合模型如下:
仿真結(jié)果如圖6~24所示,與文獻(xiàn)[17]的魯棒反演控制(Robust Back-Stepping Control,RBC)結(jié)果進(jìn)行對比。
由圖6~9可見,采用PPC時(shí)的速度跟蹤誤差與高度跟蹤誤差均被限定在預(yù)設(shè)區(qū)域內(nèi);與RBC相比,PPC能使速度跟蹤誤差與高度跟蹤誤差具有更好的動(dòng)態(tài)性能;當(dāng)參數(shù)存在攝動(dòng)時(shí),PPC的控制精度也更高,表現(xiàn)出更強(qiáng)的魯棒性。 圖10~11表明,兩種控制方法的航跡角控制效果相當(dāng),但PPC的航跡角響應(yīng)更平滑。圖12~15表明,雖然RBC的俯仰角與俯仰角速度響應(yīng)更平滑,但采用PPC時(shí),這兩個(gè)角度響應(yīng)沒有高頻抖振現(xiàn)象,并且PPC能夠保證俯仰角跟蹤誤差與俯仰角速度跟蹤誤差具有更好的動(dòng)態(tài)性能與穩(wěn)態(tài)精度。由圖16~19可見,兩種控制
方法的彈性狀態(tài)與控制輸入均沒有高頻抖振現(xiàn)象。圖20~24表明,ε0(t),ε1(t),ε2(t),ε3(t),ε4(t)均有界。由圖7,9,11,13,15可見,所有跟蹤誤差均被限定在預(yù)設(shè)包絡(luò)內(nèi),使得跟蹤誤差不僅動(dòng)態(tài)性能較好,穩(wěn)態(tài)精度同樣較高,同時(shí)控制魯棒性也滿足要求。
4 結(jié)? 論
通過設(shè)計(jì)一種新型性能函數(shù),使控制律擺脫了對跟蹤誤差精確初值的依賴,且能使所有跟蹤誤差具有良好的動(dòng)態(tài)性能與穩(wěn)態(tài)精度。所設(shè)計(jì)的控制律不含有任何自適應(yīng)參數(shù)或?qū)W習(xí)參數(shù),具有良好的實(shí)時(shí)性。同時(shí)性能函數(shù)構(gòu)型簡單,無需對虛擬控制律求導(dǎo),避免了微分項(xiàng)膨脹。仿真結(jié)果表明,所提方法能夠?qū)⒏櫿`差約束在預(yù)設(shè)的包絡(luò)內(nèi),以便于保證跟蹤誤差滿意的動(dòng)態(tài)性能與穩(wěn)態(tài)性能。
參考文獻(xiàn):
[1] 梁曉庚,田宏亮. 臨近空間高超聲速飛行器發(fā)展現(xiàn)狀及其防御問題分析[J]. 航空兵器,2016(4): 3-10.
Liang Xiaogeng,Tian Hongliang. Analysis of the Development Status and the Defense Problem of Near Space Hypersonic Vehicle[J]. Aero Weaponry,2016(4): 3-10.(in Chinese)
[2] 卜祥偉. 高超聲速飛行器控制研究進(jìn)展[J]. 航空兵器,2018(1): 47-61.
Bu Xiangwei. Progress on Flight Control of Hypersonic Flight Vehicles[J]. Aero Weaponry,2018(1): 47-61.(in Chinese)
[3] 王鵬飛,王潔,時(shí)建明,等. 吸氣式高超聲速飛行器控制研究綜述[J]. 航空兵器,2015(3): 3-7.
Wang Pengfei,Wang Jie,Shi Jianming,et al. Research Progress on Control System of Air-Breathing Hypersonic Flight Vehicles[J]. Aero Weaponry,2015(3): 3-7.(in Chinese)
[4] 賈二勇,袁濤,王鵬飛,等. 吸氣式高超聲速飛行器反演自適應(yīng)控制[J]. 航空兵器,2017(4): 14-20.
Jia Eryong,Yuan Tao,Wang Pengfei,et al. Adaptive Back-Stepping Controller Design for Air-Breathing Hypersonic Vehicle[J]. Aero Weaponry,2017(4): 14-20.(in Chinese)
[5] 李小兵,趙思源,卜祥偉,等. 高超聲速飛行器預(yù)設(shè)性能反演控制方法設(shè)計(jì)[J]. 北京航空航天大學(xué)學(xué)報(bào),2019,45(4): 650-661.
Li Xiaobing,Zhao Siyuan,Bu Xiangwei,et al. Design of Prescribed Performance Backstepping Control Method for Hypersonic Flight Vehicles[J]. Journal of Beijing University of Aeronautics and Astronautics,2019,45(4): 650-661.(in Chinese)
[6] 趙賀偉,胡云安,楊秀霞,等. 考慮彈性振動(dòng)的高超聲速飛行器預(yù)設(shè)性能控制[J]. 振動(dòng)與沖擊,2017,36(7): 241-247.
Zhao Hewei,Hu Yunan,Yang Xiuxia,et al. Prescribed Performance Control for a Hypersonic Vehicle Considering Elastic Vibration[J]. Journal of Vibration and Shock,2017,36(7): 241-247.(in Chinese)
[7] 王鵬飛,羅暢,劉承君,等.吸氣式高超聲速飛行器預(yù)設(shè)反演魯棒控制[J].彈箭與制導(dǎo)學(xué)報(bào),2021(1): 124-129.
Wang Pengfei,Luo Chang,Liu Chengjun,et al. Back-Stepping Robust Control for Air-Breathing Hypersonic Vehicle with Prescribed Performance[J]. Journal of Projectiles,Rockets,Missiles,and Guidance,2021(1): 124-129. (in Chinese)
[8] 馮振欣,郭建國,周軍. 高超聲速飛行器新型預(yù)設(shè)性能控制器設(shè)計(jì)[J]. 宇航學(xué)報(bào),2018,39(6): 656-663.
Feng Zhenxin,Guo Jianguo,Zhou Jun. Novel Prescribed Performance Controller Design for a Hypersonic Vehicle[J]. Journal of Astronautics,2018,39(6): 656-663.(in Chinese)
[9] Bolender M A,Doman D B. Nonlinear Longitudinal Dynamical Model of an Air-Breathing Hypersonic Vehicle[J]. Journal of Spacecraft and Rockets,2007,44(2): 374-387.
[10] Parker J T,Serrani A,Yurkovich S,et al. Control-Oriented Modeling of an Air-Breathing Hypersonic Vehicle[J]. Journal of Guidance,Control,and Dynamics,2007,30(3): 856-869.
[11] Fiorentini L. Nolinear Adaptive Controller Design for Air-Breathing Hypersonic Vechicles[D]. Columbus: The Ohio State University,2010.
[12] Bolender M A,Doman D B. Nonlinear Longitudinal Dynamical Model of an Air-Breathing Hypersonic Vehicle[J]. Journal of Spacecraft and Rockets,2007,44(2): 374-387.
[13] Parker J T,Serrani A,Yurkovich S,et al. Control-Oriented Modeling of an Air-Breathing Hypersonic Vehicle[J]. Journal of Guidance,Control,and Dynamics,2007,30(3): 856-869.
[14] 胡云安,耿寶亮,蓋俊峰. 初始誤差未知的不確定系統(tǒng)預(yù)設(shè)性能反演控制[J]. 華中科技大學(xué)學(xué)報(bào): 自然科學(xué)版,2014,42(8): 43-47.
Hu Yunan,Geng Baoliang,Gai Junfeng.? Prescribed Performance Backstepping Control for Uncertain Systems with Unknown Initial Errors[J]. Journal of Huazhong University of Science and Technology: Natural Science Edition,2014,42(8): 43-47.(in Chinese)
[15] 卜祥偉,吳曉燕,白瑞陽,等. 基于滑模微分器的吸氣式高超聲速飛行器魯棒反演控制[J]. 固體火箭技術(shù),2015,38(1): 12-17.
Bu Xiangwei,Wu Xiaoyan,Bai Ruiyang,et al. Sliding-Mode-Differentiator-Based Robust Backstepping Control of Air-Breathing Hypersonic Vehicles[J]. Journal of Solid Rocket Technology,2015,38(1): 12-17.(in Chinese)
[16] 李小兵,趙思源,卜祥偉,等. 高超聲速飛行器非仿射模型的預(yù)設(shè)性能反演控制[J]. 控制理論與應(yīng)用,2019,36(10): 1672-1681.
Li Xiaobing,Zhao Siyuan,Bu Xiangwei,et al. Prescribed Performance Control Method for Non-Affine Model of Hypersonic Vehicles[J]. Control Theory & Applications,2019,36(10): 1672-1681.(in Chinese)
[17] Bu X W,Wu X Y,Zhang R,et al. Tracking Differentiator Design for the Robust Backstepping Control of a Flexible Air-Breathing Hypersonic Vehicle[J]. Journal of the Franklin Institute,2015,352(4): 1739-1765.
Low Complexity Prescribed Performance Back-Stepping
Control of Waverider Vehicle
Jiang Baoxu1,Bu Xiangwei2*,Qi Qiang1
(1. College of Graduate,Air Force Engineering University,Xian 710051,China;
2. Air and Missile Defense College,Air Force Engineering University,Xian 710051,China)
Abstract: In view of the complex structure of neural network prescribed performance control method and low control real-time performance,a low-complexity prescribed performance inversion control method is proposed for waverider vehicle(WV). Firstly,a new type of performance function with simple form and independent of initial error is designed to envelop the tracking error. Secondly,using a conversion strategy,the inequality constraint on the tracking error is equivalently converted into an equality constraint by introducing a conversion error. Then,based on the conversion error,a new type of inversion controller without any learning parameters is designed for WV,and it avoids the “differential term expansion” problem of traditional back-stepping control. Finally,the simulation results show that the proposed method is expected to guarantee good dynamic performance and steady-state accuracy of tracking error.
Key words: waverider vehicle;prescribed performance;new performance function;new back-stepping controller;conversion error
收稿日期:2021-04-23
基金項(xiàng)目:中國科協(xié)青年人才托舉工程項(xiàng)目(18-JCJQ-QT-007)
作者簡介:姜寶續(xù)(1998-),男,遼寧大連人,碩士,研究方向?yàn)轱w行器控制技術(shù)。
通訊作者:卜祥偉(1987-),男,河南夏邑人,副教授,博士,研究方向?yàn)轱w行器先進(jìn)控制技術(shù)。