2023年全國碩士研究生考試考研英語一試題真題(含答案詳解+作文范文)_第1頁
已閱讀1頁,還剩33頁未讀, 繼續(xù)免費(fèi)閱讀

下載本文檔

版權(quán)說明:本文檔由用戶提供并上傳,收益歸屬內(nèi)容提供方,若內(nèi)容存在侵權(quán),請進(jìn)行舉報(bào)或認(rèn)領(lǐng)

文檔簡介

1、<p>  中文5235字,3180單詞,16500英文字符</p><p>  2012屆本科畢業(yè)設(shè)計(jì)(論文)外文</p><p><b>  文獻(xiàn)翻譯</b></p><p>  學(xué) 院 電氣與自動(dòng)化工程學(xué)院 </p><p>  專 業(yè): 自動(dòng)化

2、 </p><p>  姓 名: </p><p>  學(xué) 號(hào): </p><p>  外文出處:Proceedings of World Academy of </p><p>  Science,Engi

3、neering and Technolo- gy Volume 29 MAY 2008 </p><p>  ISSN 1307-6884 </p><p>  附 件:1.外文資料翻譯譯文;2.外文原文</p><p>  附件1:外文資料翻譯論文</p><p>  兩輪倒立擺移動(dòng)機(jī)

4、器人的實(shí)時(shí)控制</p><p>  S. W. Nawawi, M. N. Ahmad, and J. H. S. Osman</p><p>  摘要:近十年來,對兩輪倒立擺移動(dòng)機(jī)器人(T-WIP)或者通常稱為平衡機(jī)器人的研究在世界各地的機(jī)器人實(shí)驗(yàn)室里獲得了迅猛的發(fā)展。本文介紹了一種這類機(jī)器人的硬件設(shè)計(jì)方案。設(shè)計(jì)目的在于開發(fā)出一種T-WIP以及相應(yīng)的MATLAB接口設(shè)計(jì),為不穩(wěn)定線性系統(tǒng)

5、的研究和教學(xué)工作提供一種靈活的嵌入式研究平臺(tái)。也將討論并解決一些問題,諸如傳感器和執(zhí)行器的選擇,信號(hào)處理單元的設(shè)計(jì),MATLAB實(shí)時(shí)庫的代碼生成以及模型的建立和控制方案的選擇。最后將會(huì)采用著名的狀態(tài)反饋控制器去測試這個(gè)系統(tǒng)的功能。</p><p>  關(guān)鍵詞:嵌入式系統(tǒng),兩輪倒立擺移動(dòng)機(jī)器人</p><p><b> ?、?引言</b></p><

6、p>  自上世紀(jì)40年代以來,倒立擺已成為自動(dòng)控制領(lǐng)域眾多課題的研究對象。本文將會(huì)描述受著名的賽格威機(jī)器人的啟發(fā)而引出的課題的發(fā)展。最近,這種機(jī)器人引起了人們廣泛的興趣[1]。這是由于它為控制系統(tǒng)設(shè)計(jì),信號(hào)處理,分布式控制系統(tǒng)以及實(shí)時(shí)控制所需考慮的問題提供了豐富的實(shí)踐機(jī)會(huì)。JOE是一個(gè)基于倒立擺原理的按比列縮小的兩輪小車模型,它由數(shù)字信號(hào)處理器控制,并附加了模擬人類駕駛員重量的負(fù)載[2]。系統(tǒng)使用了一種利用陀螺儀傳感器及馬達(dá)編碼器

7、感知信息的線性空間控制器。利用這項(xiàng)技術(shù)設(shè)計(jì)的小車可以供人們在一個(gè)小的范圍或者工廠內(nèi)短距離使用,而不是使用污染嚴(yán)重的轎車和越野車[3]。</p><p>  這種機(jī)器人已經(jīng)成功實(shí)現(xiàn)了二維平面上的軌跡控制[4]。本文推薦的軌跡控制算法可以讓機(jī)器人自主移動(dòng),雖然很慢。這項(xiàng)研究可以被進(jìn)一步擴(kuò)展到一維平面系統(tǒng),在該系統(tǒng)中機(jī)器人具有相同的兩輪結(jié)構(gòu)及相同的數(shù)學(xué)模型,同時(shí)輸入信號(hào)也相似[5]。在成功的研究成果上,已經(jīng)報(bào)道出現(xiàn)了利

8、用LQR控制器的類似研究[5,6]。這種兩輪倒立擺的動(dòng)態(tài)模型也被從可控性及反饋線性化的角度來分析[7],要對系統(tǒng)復(fù)雜的輔助條件進(jìn)行檢查并發(fā)現(xiàn)系統(tǒng)的最大相對程度。這樣就通過了關(guān)于T-WIP運(yùn)動(dòng)控制的一般性討論[8]。</p><p>  姿態(tài)的檢測很難通過簡單地使用傳感器來實(shí)現(xiàn)。陀螺儀信號(hào)需要進(jìn)行積分運(yùn)算,但是溫漂問題不能忽視??梢允褂孟到y(tǒng)內(nèi)部動(dòng)態(tài)觀測器來解決這個(gè)問題[9]。本文提出了將自適應(yīng)算法應(yīng)用于機(jī)器人姿態(tài)的

9、預(yù)估和控制[10]。但是這個(gè)算法還沒有考慮到二維平面,而且機(jī)器人還是由地面計(jì)算機(jī)連線控制而沒有實(shí)現(xiàn)自主控制。本文提出了一種兩輪倒立擺機(jī)器人的變結(jié)構(gòu)控制,用于對目標(biāo)軌跡的跟蹤[11]工作的目的是使機(jī)器人能夠支撐自身軀干而不會(huì)從連鑄機(jī)上掉落。本文從傾角控制的角度[12]研究了保持機(jī)器人穩(wěn)定的陀螺儀以及倒立擺系統(tǒng)[13]。這項(xiàng)研究仍然對新的控制器的發(fā)展有意義。同時(shí)提出了一種常識(shí)性的保持穩(wěn)定的方法,并將這一方法應(yīng)用于未知地形的載人試驗(yàn)[14]。

10、</p><p>  本文剩余章節(jié)安排如下:第二節(jié)主要介紹硬件設(shè)計(jì),尤其是傳感器及控制算法的選擇。第三節(jié)介紹了機(jī)器人的數(shù)學(xué)建模。第四節(jié)是關(guān)于控制器的設(shè)計(jì)以及仿真結(jié)果。第五節(jié)介紹MATLAB接口配置。最后一節(jié)則是對控制器實(shí)時(shí)控制效果的驗(yàn)證及顯示,并將對實(shí)時(shí)執(zhí)行的結(jié)果進(jìn)行詳細(xì)說明。</p><p>  圖1 T-WIP移動(dòng)機(jī)器人原型機(jī)</p><p><b>

11、 ?、?硬件設(shè)計(jì)</b></p><p>  圖1展示的是T-WIP移動(dòng)機(jī)器人原型機(jī)。該機(jī)器人配備有兩個(gè)伺服驅(qū)動(dòng)器用來驅(qū)動(dòng)電機(jī),一個(gè)陀螺儀用于測量角度及擺體角速度,以及一組測速用編碼器。信號(hào)處理及控制共使用了三個(gè)微處理器,其中兩個(gè)用于伺服驅(qū)動(dòng)器的控制另外一個(gè)則用于機(jī)器人穩(wěn)定控制。雖然這種分布式的布局有利于分層控制的設(shè)計(jì),但是考慮到各個(gè)處理器的通信,這種設(shè)計(jì)還是比較復(fù)雜的。這種T-WIP移動(dòng)機(jī)器人主體是

12、由負(fù)載底盤和安裝了行星減速器的直流電機(jī)構(gòu)成,還搭載了包括作為控制器的DSP控制板,電機(jī)功率放大器,以及各種測量車身狀態(tài)的必要傳感器。電池使用螺絲固定在機(jī)器人外殼內(nèi)部,電池重量占機(jī)器人總重量的30%。機(jī)器人的輪子直接耦合到變速箱的動(dòng)力輸出軸。機(jī)器人利用扭矩CR和CL分別實(shí)現(xiàn)左右輪的控制。圖2顯示的是控制系統(tǒng)結(jié)構(gòu)框圖。</p><p><b>  圖2 控制系統(tǒng)框圖</b></p>

13、<p>  控制器采用嵌入式DSP控制板[15]。是一個(gè)基于X86和PC104總線的獨(dú)立運(yùn)動(dòng)控制器。運(yùn)動(dòng)控制卡,接口端子板集中在一個(gè)板卡中,因此具有尺寸小,接線少,實(shí)時(shí)性好及可靠性高的優(yōu)點(diǎn)。該控制器易于安裝,維護(hù),升級(jí),因此可以大大提高機(jī)器人的可靠性,也使得機(jī)器人可以在潮濕,粉塵以及震動(dòng)的惡劣工業(yè)環(huán)境使用。</p><p>  傳統(tǒng)的傾角計(jì)以及傾角傳感器響應(yīng)比較慢,并不適用于動(dòng)態(tài)角的跟蹤。另一方面,角

14、速度傳感器可以用于測量快速旋轉(zhuǎn),但是隨著時(shí)間的推移,角速度傳感器測得的值會(huì)顯著地受到溫漂,以及誤差累計(jì)的影響。雖然慣性測量單元(IMU)可以克服這些不足之處,但其體積比較大而且價(jià)格十分昂貴。因此本系統(tǒng)采用MicroStrain公司的FAS-G傳感器作為陀螺儀[16]。</p><p>  FAS-G微機(jī)電(MEM)傳感器,內(nèi)部由兩個(gè)具有低通濾波器的加速度計(jì)及壓電陶瓷陀螺儀構(gòu)成的。將角速度信號(hào)進(jìn)行時(shí)間的積分并將結(jié)果

15、與加速度計(jì)的信號(hào)相比從而消除溫漂影響。陀螺儀的輸出信號(hào)是與傾角對應(yīng)的0到5V之間的模擬電壓。通過數(shù)據(jù)采集卡讀取陀螺儀輸出信號(hào)并通過PC104總線將信號(hào)傳給PC。計(jì)算得出ADC最小分辨率對應(yīng)的傾角為0.08789度。次要角速度信號(hào)也通過軟件計(jì)算得出。圖3顯示的是二維平面上兩輪倒立擺移動(dòng)機(jī)器人的自轉(zhuǎn)角度。</p><p><b>  圖3自轉(zhuǎn)角度</b></p><p>

16、  這兩個(gè)齒輪伺服電機(jī)需要提供非常高的扭矩。為了滿足條件,系統(tǒng)采用IPM100作為電機(jī)驅(qū)動(dòng)。IPM100是一種基于DSP控制技術(shù)的全數(shù)字化智能伺服驅(qū)動(dòng)器,電壓和電流可達(dá)到36V,3A。此驅(qū)動(dòng)器內(nèi)嵌Technosoft公司高級(jí)運(yùn)動(dòng)語言(TML),因此為無刷直流電機(jī)的單軸及多軸應(yīng)用提供了一種靈活,結(jié)構(gòu)緊湊,易于實(shí)現(xiàn)的解決方案。如圖4所示,本系統(tǒng)將兩個(gè)增量式編碼器和陀螺儀連接在一起,以實(shí)現(xiàn)提供T-WIP機(jī)器人運(yùn)動(dòng)狀態(tài)信息的目的。所有的接口都是

17、基于圖2所示嵌入式系統(tǒng)的控制結(jié)構(gòu)。直線位置,速度以及偏航角和角速率可以由重力方向的兩個(gè)輪子(θRR和θRL)的角度旋轉(zhuǎn)而決定。俯仰角,θRP,機(jī)身角度,θRPL,θRPR這些角度的關(guān)系可以參考圖3。嵌入式控制器的任務(wù)是檢測增量式編碼器的反饋。然后處理反饋信息,確保T-WIP機(jī)器人在它的平衡點(diǎn)保持平衡。嵌入式控制器的命令采用IPC,這是一種MATLAB實(shí)時(shí)庫中的C語言接口。IPC可以在線運(yùn)行或者根據(jù)系統(tǒng)輸出響應(yīng)要求同時(shí)顯示實(shí)時(shí)結(jié)果。<

18、;/p><p>  圖4 傳感器信號(hào)與嵌入式DSP板的接口</p><p><b> ?、?數(shù)學(xué)建模</b></p><p>  平衡機(jī)器人的動(dòng)態(tài)性能取決于控制算法的效率以及系統(tǒng)的動(dòng)態(tài)模型[17]。在圖5所顯示的坐標(biāo)系中,對機(jī)器人進(jìn)行受力分析可以得出:T-WIP移動(dòng)機(jī)器人的動(dòng)力學(xué)特性可以由下面的運(yùn)動(dòng)方程(1-15)描述[2,8]。機(jī)器人的力學(xué)坐標(biāo)系

19、如圖5。</p><p>  圖5 T-WIP的力學(xué)坐標(biāo)系</p><p>  對于左邊的輪子(右邊的輪子也類似)有以下方程:</p><p><b>  (1)</b></p><p><b> ?。?)</b></p><p><b> ?。?)</b&g

20、t;</p><p><b> ?。?)</b></p><p><b> ?。?)</b></p><p><b> ?。?)</b></p><p><b> ?。?)</b></p><p>  機(jī)器人底盤運(yùn)動(dòng)方程:</

21、p><p><b> ?。?)</b></p><p><b> ?。?)</b></p><p><b>  (10)</b></p><p><b> ?。?1)</b></p><p>  HTL,HTR,HL,HR,VTL,VT

22、R,VL,VR代表不同自由機(jī)構(gòu)之間的反作用力。機(jī)器人參數(shù)如表Ⅰ。</p><p><b>  表Ⅰ</b></p><p>  T-WIP機(jī)器人參數(shù)</p><p>  符號(hào) 參數(shù) 值(單位)</p><p>  Xr

23、 直線位置 [m]</p><p>  θR 俯仰角 [rad]</p><p>  δ 偏航角 [rad]</p>

24、<p>  JRL,JRR Z軸瞬態(tài)轉(zhuǎn)矩 [kgm2] </p><p>  Mr 連接到左右輪的旋轉(zhuǎn)質(zhì)量(MRR=MRL=Mr) 0.420[kg]</p><p>  Jp 底盤在Z軸的轉(zhuǎn)動(dòng)慣量 0.

25、28[kgm2]</p><p>  Jδ 底盤在Y軸的轉(zhuǎn)動(dòng)慣量 1.12[kgm2]</p><p>  Mp 車身質(zhì)量 21.0[kg]</p><p>  R 車輪半徑

26、 0.106[m]</p><p>  L Z軸軸距及車輛重心 0.4[m]</p><p>  D 車輪接觸片之間的橫向距離    0.4[m]</p><p>  Yr 車輪

27、與Y軸之間的轉(zhuǎn)向位置 </p><p>  Xp 車身與X軸之間的轉(zhuǎn)向位置</p><p>  g 重力加速度 9.8[ms-2]</p><p>  CL,CR 左右輪的輸入轉(zhuǎn)矩</p

28、><p>  方程(1)-(11)可以利用狀態(tài)空間表示:</p><p><b> ?。?2)</b></p><p>  其中及分別作用于狀態(tài)和控制。是非線性動(dòng)態(tài)函數(shù)矩陣,是非線性輸入函數(shù)矩陣。狀態(tài)x在系統(tǒng)的定義為:</p><p><b> ?。?3)</b></p><p>

29、;  修改以上方程,然后將周圍工作點(diǎn)()進(jìn)行線性化處理并去耦。系統(tǒng)的狀態(tài)空間方程可以寫成矩陣形式:</p><p> ?。?4) </p><p><b>  (15)</b></p><p><b>  這里:<

30、;/b></p><p>  這里為了簡單起見,方程(14)和方程(15)就不再詳細(xì)描述,具體詳見文章他處[19]。通過數(shù)學(xué)建模得出的T-WIP機(jī)器人平衡模型,即方程(14)將應(yīng)用于機(jī)器人的平衡控制中。</p><p><b> ?、?控制器設(shè)計(jì)</b></p><p>  系統(tǒng)性能(即抗干擾能力以及跟蹤驅(qū)動(dòng)器輸入等)由極點(diǎn)配置控制器決定

31、。為了測試T-WIP機(jī)器人的性能,本文應(yīng)用了兩級(jí)不同的極點(diǎn)配置控制器。對于一個(gè)選擇了的極點(diǎn),控制的增益計(jì)算和實(shí)現(xiàn)在嵌入式主板上進(jìn)行。而后進(jìn)行了T-WIP機(jī)器人的配置及響應(yīng)測試,數(shù)據(jù)由控制系統(tǒng)記錄。其中包括一項(xiàng)測試,對多于一個(gè)重心的機(jī)器人進(jìn)行脈沖干擾測試,結(jié)果表明,一個(gè)擺動(dòng)傳送的能量大約為1.2J。</p><p>  在記錄的響應(yīng)上可以清楚的解決阻尼比和穩(wěn)定時(shí)間等問題并進(jìn)行有效的系統(tǒng)微調(diào)。圖6顯示的是上述測試的系

32、統(tǒng)響應(yīng),此時(shí)系統(tǒng)最初選擇的極點(diǎn)為[-1.5,-1.5,-0.5-3i,-0.5+3i]。要特別注意當(dāng)系統(tǒng)阻尼很小的時(shí)候會(huì)出現(xiàn)很大的振蕩。圖7給出的是當(dāng)極點(diǎn)改為[-1.5-i,-1.5+i,-3.5-5i,-3.5+5i]時(shí)增加系統(tǒng)阻尼比的結(jié)果。由此可以看出,系統(tǒng)響應(yīng)得到顯著提高?,F(xiàn)在,系統(tǒng)有一個(gè)微弱的干擾力。當(dāng)干擾力作用于T-WIP機(jī)器人,會(huì)導(dǎo)致倒立擺向前倒()。</p><p>  控制系統(tǒng)加快車輪速度使機(jī)器人

33、在向積極的方向前進(jìn),最終使擺錘落在另一個(gè)方向。然后利用負(fù)轉(zhuǎn)矩,使車輛移動(dòng)到原來的位置并保持?jǐn)[錘直立??刂破鞯娜蝿?wù)是確保:</p><p>  對于一個(gè)給定的 (16)</p><p>  這是一個(gè)物理意義上的問題,因?yàn)槭褂眠@些參考命令可以安全的遵循運(yùn)動(dòng)計(jì)劃?;鞠敕ㄊ抢酶┭鼋铅茸鳛檐嚨挠烷T實(shí)現(xiàn)加減速,直到達(dá)到指定速度。</p><p>  圖6

34、 倒立擺系統(tǒng)初始化極點(diǎn)位置及作用于擺錘的脈沖干擾力的響應(yīng)</p><p> ?。芰總鬏敗?.2J)</p><p>  圖7 改進(jìn)極點(diǎn)位置倒立擺系統(tǒng)及作用于擺錘的脈沖干擾力的響應(yīng)</p><p>  (能量傳輸≈1.2J)</p><p>  另外,在測試中解決了駕駛性能的問題。為了成功地提高駕駛性能,機(jī)器人以兩項(xiàng)條件為標(biāo)準(zhǔn)。第一個(gè)條件是解

35、析系統(tǒng)對漸變速度輸入的響應(yīng),第二個(gè)條件是讓不同的駕駛者感受T-WIP的操作。</p><p>  將不同駕駛者的感受與系統(tǒng)行為的解析相結(jié)合讓T-WIP控制系統(tǒng)的進(jìn)一步改進(jìn)得到可能。圖8顯示的是最終極點(diǎn)選擇后系統(tǒng)對快速斜坡輸入的響應(yīng)。請注意,最大加速度可能低于最大減速。由于電機(jī)的速度-電流特性高轉(zhuǎn)矩不能得到高速度,然而這正說明機(jī)器人必須在加速階段的最后才能回到直立位置。減速需要低速狀態(tài)下的最大轉(zhuǎn)矩,因而可能出現(xiàn)一個(gè)

36、急劇的減速。</p><p>  可以將兩極點(diǎn)向左移動(dòng)從而實(shí)現(xiàn)通過移動(dòng)極點(diǎn)而進(jìn)一步提高系統(tǒng)性能的目的,也因而可以使系統(tǒng)速度更快[20]。尺側(cè)間隙以及最大轉(zhuǎn)矩可以傳送到地面(抓地力)在一定限度內(nèi)可以防止協(xié)調(diào)器移動(dòng)極點(diǎn)。使用一種非線性極點(diǎn)自適應(yīng)控制器(取決于系統(tǒng)狀態(tài))將使系統(tǒng)性能進(jìn)一步提高。</p><p>  圖8 斜坡形速度輸入的響應(yīng)</p><p> ?、?MAT

37、LAB接口設(shè)計(jì)</p><p>  本系統(tǒng)中的嵌入式控制系統(tǒng)基于MATLAB實(shí)時(shí)庫,所以嵌入式控制器卡與MATLAB之間的接口協(xié)議非常重要。T-WIP的實(shí)時(shí)平衡控制將在MATLAB 實(shí)時(shí)庫中全面展開。因此,它具有一個(gè)優(yōu)勢就是可以智能化的觀測實(shí)時(shí)結(jié)果,并檢測集成執(zhí)行器后控制器的性能。此外可以很方便的對控制器進(jìn)行不斷優(yōu)化直到用戶滿意。</p><p>  在MATLAB實(shí)時(shí)庫中,特殊的實(shí)時(shí)內(nèi)核

38、模型取代windows消息響應(yīng)機(jī)制。因此,實(shí)時(shí)模型的性能對于獲得更好的系統(tǒng)響應(yīng)已經(jīng)足夠好了。</p><p>  圖9 MATLAB RTW內(nèi)核模型</p><p>  RTW是以Simulink仿真圖為原型,以在各種目標(biāo)計(jì)算機(jī)平臺(tái)上進(jìn)行測試,配置實(shí)時(shí)系統(tǒng)為應(yīng)用目的而建立的。利用實(shí)時(shí)庫,用戶可直接生成源代碼,并且這些代碼可以包含編譯器,輸入輸出設(shè)備,內(nèi)存模型,通信方式以及用戶的應(yīng)用程序所需

39、的其他特性。</p><p>  配置設(shè)置的第一步是安裝包含實(shí)時(shí)視窗對象的MATLAB以及Visual C/C++軟件。再通過在MATLBA中輸入相關(guān)命令激活實(shí)時(shí)視窗對象內(nèi)核,然后通過菜單選擇MATLAB中的C編譯器。下面將介紹一種基于MATLAB RTW的傳感器采樣方法。在設(shè)計(jì)一個(gè)用C語言寫的子函數(shù)時(shí),應(yīng)該定義該函數(shù)的模數(shù)轉(zhuǎn)換通道及輸出參數(shù)指標(biāo)。GetAD.c的部分代碼如下:</p><p&

40、gt;  #define S_FUNCTION_NAME GetAD</p><p>  #define NUM_PARAS (1)</p><p>  #define AD_CHANNEL _PARAM(ssGetSFcnParam(S,0))</p><p>  #define AD_CHANNEL (real_T)mxGetPr(AD_CHANNEL_P

41、ARAM)[0]</p><p>  在GetAD.c代碼在MATLAB中編譯沒有錯(cuò)誤后,最終生成子函數(shù)。然后模數(shù)轉(zhuǎn)換的參數(shù)將會(huì)配置到通道1。其次還需要對另一個(gè)系統(tǒng)目標(biāo)文件GetAD.tlc進(jìn)行設(shè)計(jì),并將其保存在GetAD.c相同目錄下。GetAD.tlc主要代碼的流程圖如下: </p><p>  圖10 GetAD.clt代碼流程圖</p><p>  目標(biāo)文件

42、GetAD.tlc中的“_outp”命令及“_inp”命令作為參考之用,用于嵌入式控制器與MATLAB之間的數(shù)據(jù)收發(fā)。GetAD.tlc文件中的ctrl_byte=0xff用于控制DA卡停止在PC104總線上查詢數(shù)據(jù)。</p><p>  然后編譯GetAD.tlc并確保不出錯(cuò)。</p><p>  在GetAD.tlc仿真參數(shù)特性中,。此配置是用于MATLAB中的實(shí)時(shí)庫的設(shè)置。采樣時(shí)間定

43、為為5ms并設(shè)置解算器選項(xiàng)及固定步長。</p><p>  在選擇為外部模式后對該文件編譯。圖11顯示的是傳感器的輸出信號(hào)。</p><p>  圖11 陀螺儀角度采樣</p><p>  設(shè)計(jì)其他子函數(shù)也可以使用同樣的方法。最后提出T-WIP的MATLBA實(shí)時(shí)庫接口?,F(xiàn)在T-WIP測試臺(tái)可以用于任何系統(tǒng)結(jié)構(gòu)相似的控制器。系統(tǒng)結(jié)構(gòu)如圖12。它主要包含三個(gè)模塊,分別是

44、輸入模塊,控制器模塊以及實(shí)時(shí)庫模塊。同樣,輸入函數(shù)可以替代速度和方向的輸入?yún)⒖肌?lt;/p><p><b> ?、?實(shí)驗(yàn)結(jié)果</b></p><p>  為了測試所開發(fā)的T-WIP硬件系統(tǒng)以及上一節(jié)提出的基于極點(diǎn)配置算法的控制器。圖13顯示的是系統(tǒng)響應(yīng),與圖6顯示的仿真結(jié)果十分吻合,這表明T-WIP在閉環(huán)系統(tǒng)中運(yùn)行的很好。從圖14可以看出,位置,速度,角速度和傾角的輸出

45、響應(yīng)曲線與模擬結(jié)果曲線相似。位置和速度的穩(wěn)態(tài)誤差幾乎為零。還表明,角速度及角度的穩(wěn)態(tài)誤差也接近零。極點(diǎn)控制器可以在穩(wěn)態(tài)誤差及校正時(shí)間方面有優(yōu)良的性能。</p><p>  圖12 采用極點(diǎn)配置控制器保持T-WIP平衡的實(shí)時(shí)控制界面</p><p>  圖13 使用極點(diǎn)配置控制器的T-WIP輸出響應(yīng)結(jié)果</p><p>  圖13顯示的是圖14放大的結(jié)果,它表明系統(tǒng)處

46、于小幅度擾動(dòng)的平衡狀態(tài)。底盤的擾動(dòng)范圍大約為4×10-3m,主干的振幅為0.005rad。</p><p>  為進(jìn)一步研究,將使用各種不同類型的非線性控制器來控制T-WIP機(jī)器人。其中,他們大都是狀態(tài)反饋線性化的滑模控制。</p><p>  圖14 采用極點(diǎn)配置控制器的T-WIP對平衡振動(dòng)的實(shí)時(shí)控制結(jié)果</p><p><b> ?、?結(jié)論&

47、lt;/b></p><p>  本文提出了一種T-WIP機(jī)器人系統(tǒng)的發(fā)展。解決了動(dòng)力學(xué)建模,執(zhí)行器及控制器的選型,基于MATLAB的接口配置和嵌入式控制器的設(shè)計(jì)以及極點(diǎn)配置控制策略的實(shí)現(xiàn)等問題。通過實(shí)驗(yàn)表明采用MATLAB實(shí)時(shí)庫的嵌入式控制系統(tǒng)可以穩(wěn)定運(yùn)行,傳感器返回的信號(hào)也比較理想。本文得出結(jié)論:基于MATLAB的控制器架構(gòu)可以提供理想的實(shí)驗(yàn)結(jié)果。</p><p><b&g

48、t;  致謝</b></p><p>  感謝Kaustubh Pathak對本文原版幫助。</p><p><b>  附錄2 外文原文</b></p><p>  Real-Time Control of a Two-Wheeled Inverted Pendulum Mobile Robot</p><p&

49、gt;  S. W. Nawawi, M. N. Ahmad, and J. H. S. Osman</p><p>  Abstract—The research on two-wheeled inverted pendulum (T-WIP) mobile robots or commonly known as balancing robots have gained momentum over the la

50、st decade in a number of robotic laboratories around the world. This paper describes the hardware design of such a robot. The objective of the design is to develop a T-WIP mobile robot as well as MATLAB interfacing confi

51、guration to be used as flexible platform comprises of embedded unstable linear plant intended for research and teaching purposes. Iss</p><p>  Keywords—Embedded System, Two-wheeled Inverted Pendulum Mobile R

52、obot.</p><p>  I. INTRODUCTION</p><p>  Inverted pendulum has been the subject of numerous studies in automatic control since the forties. In this paper, the development of the theme inspired by

53、 the well known Segway robot is described. This kind of robots has induced a lot of interest recently [1]. This is due to the fact that it provides rich opportunities for application of control design, signal processing,

54、 distributed control systems and consideration of real time constraint for implementation issues. A scaled down prototype of a</p><p>  The trajectory control for this type of robot was successfully implemen

55、ted in two dimensional by [4]. The proposed trajectory control algorithm can make the robot move autonomously, although quite slowly. The work was further extended in [5] where the robot is assumed to receive similar inp

56、ut on both wheels and the mathematical modeling is represented in one dimensional plane system. Based on successful result in [5], similar work has been reported using LQR controller [6]. The dynamic model of </p>

57、<p>  The attitude is difficult to detect by simply using the signals from sensor. The integration of rate gyroscope signal has contributed the problem of drift. The problem solve using an observer that considering

58、 internal dynamic of the system [9]. The estimation and control algorithm of the posture using the adaptive observer is proposed by [10]. The presented algorithm did not consider movement on the two dimensional plane and

59、 the robot also was not autonomous and connected by wires with ground co</p><p>  The remaining of the paper is organized as follows: Section II treats on hardware design , particularly in the selection of s

60、ensors and their associated algorithms Section III describes the mathematical modeling of the robot. Section IV addressed the controller design and their simulation results. Section V presents the MATLAB interfacing conf

61、iguration. Finally the verification of the controller using real time implementation is shown experimentally. Results of the real time implementation will </p><p>  II. HARDWARE DEVELOPMENT</p><p&

62、gt;  Fig. 1 shows the prototype design of the robot. The robot is equipped with two servo drives for actuation, a Gyroscope for measuring angle and angular velocity of pendulum body, and encoders for measuring the positi

63、on of the wheels. Signal processing and control algorithm are distributed among the three microprocessors. Two of them are used for servo drives while the other one is used for stabilizing control. Although this kind of

64、layout enables hierarchical control design, it also complicates i</p><p>  Fig. 1 The prototype T-WIP mobile robot</p><p>  The T-WIP mobile robot is composed of a chassis carrying a DC motor co

65、upled to a planetary gearbox for each wheel, the DSP board used to implement the controller, the power amplifiers for the motors, the necessary sensors to measure the vehicle’s states. The battery is bolted inside the bo

66、dy casing and it significantly represents 30% of the total robot mass. The wheels of the vehicle are directly coupled to the output shaft of the gearboxes. The robot is control by applying a torque CR and CL to</p>

67、<p>  The controller is implemented on an Embedded DSP board [15]. It is a standalone motion controller based on combination of embedded PC104 main board of X86, motion control board, terminal board in one structu

68、re, and thus has the advantages of smaller dimension, less wiring, real time capability and higher reliability. It is easy to upgrade, install and maintain, and thus increase the reliability of the robot to operate under

69、 adverse industrial environments, such as humid, dust, and vibration.</p><p>  Conventional inclinometers, or analog tilt sensors, typically exhibit slow response and cannot be used to track dynamic angular

70、motion. On the other hand, angular rate sensors can be used to measure fast rotations, but they suffer from significant drift and error accumulation over time. Inertial measurement units (IMU’s)</p><p>  Fig

71、.2 Control Architecture</p><p>  can be used to overcome these limitations, but these are relatively large and expensive. As such, the FAS-G sensor from MicroStrain is used as the gyro sensor [16].</p>

72、<p>  Employing micro-electromechanical (MEM) sensors, FAS-G consists of a combination of two low pass filtered accelerometers and one piezo-ceramic gyro. The angular rate signal is integrated internally over time

73、 and compared to the accelerometer signal to eliminate drift. The gyro output signal is an analog voltage between 0 and 5 volts corresponding to the angle of tilt. This signal is read from the Data Acquisition Card and t

74、he result is passed to PC by PC104 data bus. It was calculated that one A</p><p>  Fig. 3 Angle of Rotational</p><p>  Both geared servo motor needs to generate a very high torque. To achieve th

75、is, the IPM100 is used as the motor driver. It is basically a 36V, 3A fully digital intelligent servo drive based on the DSP controller technology. It is also embedded with the high level Technosoft Motion Language (TML)

76、 and therefore offers a flexible, compact and easy to implement solution for single or multi-axis applications with brushless and DC motors.</p><p>  To provide information about T-WIP states for control pur

77、poses, two incremental encoders and a rate gyroscope are interfaced together as shown in Fig. 4. All the interfacing is based on control structure of embedded system seen in Fig. 2 above. Straight line position and speed

78、 as wells as yaw angle and rate can be determined from the angle rotation of the two wheels (θRR and θRR) with respect to the gravity. The relation of these angles with pitch angle, θP and the body angle, θRPL and θRPR c

79、an </p><p>  Fig. 4 Interface between sensor signals and the Embedded DSP board</p><p>  III. MATHEMATICAL MODELING</p><p>  The dynamic performance of a balancing robot depends on

80、the efficiency of the control algorithms and the dynamic model of the system [17]. By adopting the coordinate system shown in Fig. 5 using Newtonian mechanics, it can be shown that the dynamics of the T-WIP mobile robot

81、under consideration is governed by the following motion equations (1-15) [2],[18]. The coordinate system for the robot is depicted in Fig. 5.</p><p>  Fig. 5 Coordinate system of the T-WIP</p><p&g

82、t;  For left hand wheel (analogous for right hand wheel):</p><p><b> ?。?)</b></p><p><b>  (2)</b></p><p><b> ?。?)</b></p><p><b&g

83、t; ?。?)</b></p><p><b>  (5)</b></p><p><b> ?。?)</b></p><p><b> ?。?)</b></p><p>  For the chassis, the equations:</p>&l

84、t;p> ?。?) </p><p> ?。?) </p><p><b> ?。?0)</b></p><p><b>  (11)</b></p><p>  where HTL, HTR, HL, HR, VTL, VTR,

85、VL, VR represent reaction forces between the different free bodies. The robot parameters are as tabulated in Table I.</p><p>  Equations (1)-(11) can be represented in the state-space form as:</p><

86、;p><b> ?。?2)</b></p><p>  where, are respectively the state and the control. is nonlinear dynamic function matrix and is nonlinear input function matrix. The state, x of the system is defined

87、as:</p><p><b>  (13)</b></p><p>  Modifying the equations above and then linearizing the result around the operating point (θp=0, xr=0, δ=0) and de-coupling,the system’s state space

88、equations can be written in matrix form as:</p><p>  (14) </p><p><b>  (15)</b></p><p><b>  Where</b&

89、gt;</p><p>  For simplicity, the details of equation (14) and (15) are not shown here and can be found elsewhere [19]. The T-WIP balancing model, namely equation (14) will be used through out this work conse

90、quently.</p><p>  IV. CONTROLLER DESIGN</p><p>  System performance (i.e. reaction to disturbance forces, tracking of driver input, etc.) is driven by the pole placement controller. In order to

91、test the T-WIP performance, pole placement controllers with different poles has been applied. </p><p>  For a chosen pole placement, the controller’s gains were calculated and implemented on the embedded boa

92、rd. T-WIP was then tested with the configuration and the response is then recorded by the control system. One of the tests conducted consist of an impulse disturbance force applied to a position above the center of gravi

93、ty. The energy transmitted with a falling weight amounted to about 1.2 J. </p><p>  Issues like damping ratio and settling time could be clearly identified on the recorded responses and permitted an efficien

94、t fine-tuning of the system. Fig. 6 shows the system’s response to the above mentioned test with the initial pole placement chosen at pole [-1.5,-1.5,-0.5-3i,-0.5+3i]. Note the pronounce oscillation of the system which i

95、ndicates too weak damping. Increasing the damping ratio when change the pole to [-1.5-i,- 1.5+i,-3.5-5i,-3.5+5i] give the result as shown in Fig. 7. It can b</p><p>  for a given (16)</p>&

96、lt;p>  This is a physically meaningful problem because using these reference commands, one can safely follow a motion plan. The essential idea to use the pitch angle θp as a gas pedal for vehicle and use it to acceler

97、ate and decelerate until the specified speed is attained.</p><p>  Another issue that has been addressed during testing is drivability. In order to successfully improve drivability, it was characterized base

98、d on two criteria. First criteria are readouts of the system’s reaction to a ramp shaped speed input and second criteria are the way different drivers felt about T-WIP handling.</p><p>  Combining the driver

99、’s feelings with the readouts of system behavior allowed further improvement of T-WIP control system. Fig. 8 shows the system’s response to a velocity ramp input with the final pole placement chosen. Note that the maximu

100、m acceleration possible is lower than the maximum deceleration. Due to the motor’s speed-current characteristics, a high torque cannot be obtained when operating at high</p><p>  speeds. However, this is exa

101、ctly what is necessary to get the vehicle back into an upright position at the end of the acceleration phase. Deceleration demands maximum torque at low speeds so a steeper ramp is therefore possible.</p><p>

102、;  Fig. 6 Initial pole placement of the “pendulum” system and associated response to an impulse disturbance force (energy transmitted ≈1.2 J) applied to the pendulum</p><p>  Fig. 7 Improved pole placement o

103、f the “pendulum” system and associated response to an impulse disturbance force (energy transmitted≈1.2 J) applied to the pendulum</p><p>  Increasing performance with the pole placement chosen can be achiev

104、ed by moving the poles further to the left, thus making the system faster [20]. Backlash as well the maximum torque that can be transmitted to the ground (grip) prevent tuners from moving the poles past a certain limit.

105、The used of an adaptive pole placement and nonlinear controller (depending on the system’s state) would enable further improvements to the system.</p><p>  Fig. 8 Reaction to a ramp shaped speed input</p&

溫馨提示

  • 1. 本站所有資源如無特殊說明,都需要本地電腦安裝OFFICE2007和PDF閱讀器。圖紙軟件為CAD,CAXA,PROE,UG,SolidWorks等.壓縮文件請下載最新的WinRAR軟件解壓。
  • 2. 本站的文檔不包含任何第三方提供的附件圖紙等,如果需要附件,請聯(lián)系上傳者。文件的所有權(quán)益歸上傳用戶所有。
  • 3. 本站RAR壓縮包中若帶圖紙,網(wǎng)頁內(nèi)容里面會(huì)有圖紙預(yù)覽,若沒有圖紙預(yù)覽就沒有圖紙。
  • 4. 未經(jīng)權(quán)益所有人同意不得將文件中的內(nèi)容挪作商業(yè)或盈利用途。
  • 5. 眾賞文庫僅提供信息存儲(chǔ)空間,僅對用戶上傳內(nèi)容的表現(xiàn)方式做保護(hù)處理,對用戶上傳分享的文檔內(nèi)容本身不做任何修改或編輯,并不能對任何下載內(nèi)容負(fù)責(zé)。
  • 6. 下載文件中如有侵權(quán)或不適當(dāng)內(nèi)容,請與我們聯(lián)系,我們立即糾正。
  • 7. 本站不保證下載資源的準(zhǔn)確性、安全性和完整性, 同時(shí)也不承擔(dān)用戶因使用這些下載資源對自己和他人造成任何形式的傷害或損失。

最新文檔

評(píng)論

0/150

提交評(píng)論