機(jī)械設(shè)計(jì)外文翻譯-動(dòng)態(tài)優(yōu)化的一種新型高速高精度的三自由度【中文4200字】【PDF+中文WORD】,中文4200字,PDF+中文WORD,機(jī)械設(shè)計(jì),外文,翻譯,動(dòng)態(tài),優(yōu)化,一種,新型,高速,高精度,自由度,中文,4200,PDF,WORD 【中文4200字】 動(dòng)態(tài)優(yōu)化的一種新型高速，高精度的三自由度機(jī)械手① 彭蘭（蘭朋）②，魯南立，孫立寧，丁傾永 (機(jī)械電子工程學(xué)院，哈爾濱理工學(xué)院，哈爾濱 150001，中國(guó)) ( Robotics Institute。Harbin Institute of Technology，Harbin 150001，P。R。China) 摘要 介紹了一種動(dòng)態(tài)優(yōu)化三自由度高速、高精度相結(jié)合，直接驅(qū)動(dòng)臂平面并聯(lián)機(jī)構(gòu)和線性驅(qū)動(dòng)器，它可以提高其剛度進(jìn)行了動(dòng)力學(xué)分析軟件ADAMS仿真模擬環(huán)境中，進(jìn)行仿真模擬實(shí)驗(yàn).設(shè)計(jì)調(diào)查是由參數(shù)分析工具完成處理的，分析了設(shè)計(jì)變量的近似的敏感性，包括影響參數(shù)的每道光束截面和相對(duì)位置的線性驅(qū)動(dòng)器上的性能.在適當(dāng)?shù)姆绞较拢Ｐ涂梢垣@得一個(gè)輕量級(jí)動(dòng)態(tài)優(yōu)化和小變形的參數(shù)。一個(gè)平面并聯(lián)機(jī)構(gòu)不同截面是用來(lái)改進(jìn)機(jī)械手的.結(jié)果發(fā)生明顯的改進(jìn)后的系統(tǒng)動(dòng)力學(xué)仿真分析和另一個(gè)未精制一個(gè)幾乎是幾乎相等.但剛度的改進(jìn)的質(zhì)量大大降低，說(shuō)明這種方法更為有效的。 關(guān)鍵詞: 機(jī)械手、ADAMS、優(yōu)化、動(dòng)力學(xué)仿真 0 簡(jiǎn)介 并聯(lián)結(jié)構(gòu)機(jī)械手（PKM）是一個(gè)很有前途的機(jī)器操作和裝配的電子裝置，因?yàn)樗麄冇幸恍┟黠@的優(yōu)勢(shì)，例如：串行機(jī)械手的高負(fù)荷承載能力，良好的動(dòng)態(tài)性能和精確定位的優(yōu)點(diǎn)等. 一種新型復(fù)合3一DOF臂的優(yōu)點(diǎn)和串行機(jī)械手，也是并聯(lián)機(jī)構(gòu)為研究對(duì)象，三自由度并聯(lián)機(jī)器人是少自由度并聯(lián)機(jī)器人的重要類型。三自由度并聯(lián)機(jī)器人由于結(jié)構(gòu)簡(jiǎn)單，控制相對(duì)容易，價(jià)格便宜等優(yōu)點(diǎn)，具有很好的應(yīng)用前景。但由于它們比六自由度并聯(lián)機(jī)器人更復(fù)雜的運(yùn)動(dòng)特性，增加了這類機(jī)構(gòu)型綜合的難度，因此對(duì)三自由度并聯(lián)機(jī)器人進(jìn)行型綜合具有理論意義和實(shí)際價(jià)值。本文利用螺旋理論對(duì)三自由度并聯(lián)機(jī)器人進(jìn)行型綜合，以總結(jié)某些規(guī)律，進(jìn)一步豐富型綜合理論，并為新機(jī)型的選型提供理論依據(jù)，以下對(duì)其進(jìn)行闡述。 如圖-1所示 機(jī)械手組成的平面并聯(lián)機(jī)構(gòu)(PPM)包括平行四邊形結(jié)構(gòu)和線性驅(qū)動(dòng)器安裝在PPM.兩直接驅(qū)動(dòng)電機(jī)c整合交流電高分辨率編碼器的一部分作為驅(qū)動(dòng)平面并聯(lián)機(jī)械裝置.線型致動(dòng)器驅(qū)動(dòng)的聲音線圈發(fā)動(dòng)機(jī).這被認(rèn)為是理想的驅(qū)動(dòng)短行程的一部分.作為一個(gè)非換直接驅(qū)動(dòng)類，音圈電機(jī)可以提供高位置敏感和完美的力量與中風(fēng)的角色，高精密線性編碼作為回饋部分保證在垂直方向可重復(fù)性。 另一方面，該產(chǎn)品具有較高的剛度比串行機(jī)械手，因?yàn)樗奶攸c(diǎn)和低封閉環(huán)慣性轉(zhuǎn)矩。同時(shí)，該系統(tǒng)可以克服了柔性耦合力學(xué)彈性、齒輪、軸承、被撕咬支持，連接軸和其他零件，包括古典驅(qū)動(dòng)設(shè)備，因此該機(jī)械手是更容易得到動(dòng)力學(xué)性能好、精度高。 圖-1 3自由度的混合結(jié)構(gòu)的機(jī)械手 當(dāng)長(zhǎng)度的各個(gè)環(huán)節(jié)的平面并聯(lián)機(jī)時(shí)，構(gòu)決定于運(yùn)動(dòng)學(xué)分析和綜合[4-7]，機(jī)械優(yōu)化設(shè)計(jì)的首要任務(wù)，應(yīng)加大僵硬、降低質(zhì)量.關(guān)于幾個(gè)參數(shù)模型.這是它重要和必要的影響，研究了各參數(shù)對(duì)模型表現(xiàn)以進(jìn)一步優(yōu)化。本文就開(kāi)展設(shè)計(jì)研究工具，通過(guò)參數(shù)分析亞當(dāng)斯，又要適當(dāng)?shù)姆绞絹?lái)獲得一個(gè)輕量級(jí)的優(yōu)化和小變形系統(tǒng)。 1 仿真模型 ADAMS(Automatic Dynamic Analysis 0f Mechanical System)自動(dòng)機(jī)械系統(tǒng)動(dòng)力學(xué)分析是一個(gè)完美的軟件，對(duì)機(jī)械系統(tǒng)動(dòng)力學(xué)模擬可處理機(jī)制包括有剛性和靈活的部分，仿真模型可以創(chuàng)造出機(jī)械手的亞當(dāng)斯環(huán)境 如圖-2。OXYz是全球性的參考幀，并OXYz局部坐標(biāo)系，兩個(gè)直流驅(qū)動(dòng)電機(jī)、交流和02M O1A表示，與線性驅(qū)動(dòng)器CH被視為剛性轉(zhuǎn)子轉(zhuǎn)動(dòng)慣量電機(jī)傳動(dòng)的120kg/cm2。大眾的線性驅(qū)動(dòng)器是1.5kg，連接AB、德、03F和LJ被視為柔性體立柱、橫梁GK，通用公司和公里，形成一個(gè)三角形，也被當(dāng)作柔性傳動(dòng)長(zhǎng)度的鏈接是決定提前運(yùn)動(dòng)學(xué)設(shè)計(jì)為AB =O3F = 7cm，DE=IJ=7cm，GK= 7cm，GM =11.66cm， = 8.338cm。其它維度，這個(gè)數(shù)字是01A = 02M =7cm，CB=CD=HJ 2.5cm。EF=EG=JK= 3cm。 雖然總平面并聯(lián)機(jī)構(gòu)的運(yùn)動(dòng)都是在水平、垂直和水平剛度必須在豎向剛度特征通常低于水平僵硬，因?yàn)樗慕巧诖怪睉冶哿旱慕孛娉叽缬?jì)算每一束平面并聯(lián)機(jī)構(gòu)和相對(duì)位置的線性驅(qū)動(dòng)器是兩個(gè)非常僵硬的影響因素的系統(tǒng)。 運(yùn)動(dòng)支鏈可分為三類："主動(dòng)鏈（由驅(qū)動(dòng)器賦予確定獨(dú)立運(yùn)動(dòng)的支鏈。一般是單驅(qū)動(dòng)器控制一個(gè)自由度的運(yùn)動(dòng)），從動(dòng)鏈（不帶驅(qū)動(dòng)器、被迫作確定運(yùn)動(dòng)的支鏈。又分為以下兩種：約束鏈：獨(dú)立限制機(jī)構(gòu)自由度的從動(dòng)鏈。冗余鏈：重復(fù)限制機(jī)構(gòu)自由度的從動(dòng)鏈）復(fù)合鏈（有單驅(qū)動(dòng)器、但限制一個(gè)以上的機(jī)構(gòu)自由度的支鏈，實(shí)際是主動(dòng)鏈與約束鏈的組合）-并聯(lián)機(jī)構(gòu)是由這幾種支鏈用不同形式組合起來(lái)的。動(dòng)鏈中的約束鏈除了可以提高機(jī)構(gòu)剛度和作為測(cè)量鏈外，其更主要的作用是用來(lái)約束動(dòng)平臺(tái)的某一個(gè)或幾個(gè)自由度，以使其實(shí)現(xiàn)預(yù)期的運(yùn)動(dòng)。 圖-2 仿真模型 2 仿真模擬結(jié)果 在本節(jié)中，平均位移的末端是用來(lái)描述動(dòng)態(tài)剛度，這是在不同的配置在不同的線性驅(qū)動(dòng)器向前，從最初的位置的目的地，一般的豎向位移的機(jī)械手是作為目標(biāo)來(lái)研究豎向剛度，平均差別的橫坐標(biāo)、縱坐標(biāo)點(diǎn)之間有一個(gè)剛性數(shù)學(xué)模型，模型，作為目標(biāo)來(lái)研究水平剛度。 并聯(lián)機(jī)器人的構(gòu)型設(shè)計(jì)即型綜合是并聯(lián)機(jī)器人設(shè)計(jì)的首要環(huán)節(jié)，其目的是在給定所需自由度和運(yùn)動(dòng)要求條件下，尋求并聯(lián)機(jī)構(gòu)桿副配置、驅(qū)動(dòng)方式和總體布局等的各種可能組合。國(guó)內(nèi)的許多學(xué)者正致力于這方面的研究，其中比較有代表性的有如下幾種方法："黃真為代表的約束綜合法；楊廷力等人的結(jié)構(gòu)綜合法；代表的李代數(shù)綜合法。以上各種方法自成體系，各有特點(diǎn)，都缺乏理論的完備性。本文提出添加約束法，是從限制自由度的角度出發(fā)，增加約束，去除不需要的自由度，因每條主動(dòng)鏈只有一個(gè)驅(qū)動(dòng)裝置，讓其控制一個(gè)自由度，其余自由度通過(guò)純約束鏈去除，這樣可以使主、從動(dòng)運(yùn)動(dòng)鏈的作用分離，運(yùn)動(dòng)解耦，有利于控制。具有三自由度的并聯(lián)機(jī)床，當(dāng)采用條主動(dòng)支鏈作為驅(qū)動(dòng)時(shí)，機(jī)構(gòu)就需要約束另三個(gè)自由度，通過(guò)選擇無(wú)驅(qū)動(dòng)裝置的從動(dòng)鏈來(lái)完成，則整個(gè)機(jī)構(gòu)成為有確定運(yùn)動(dòng)的三自由度的并聯(lián)機(jī)構(gòu)。黃真等提出的約束綜合法對(duì)完全對(duì)稱的少自由度并聯(lián)機(jī)器人機(jī)構(gòu)進(jìn)行了型綜合，完全對(duì)稱的支鏈結(jié)構(gòu)相同，都屬于復(fù)合鏈，每條支鏈除都有一個(gè)單驅(qū)動(dòng)器，控制一個(gè)自由度外，還應(yīng)約束一個(gè)以上自由度才能使機(jī)構(gòu)的六個(gè)自由度全部受控，使機(jī)構(gòu)有確定的運(yùn)動(dòng)。 2.1 截面效應(yīng) 扭轉(zhuǎn)變形位移的連結(jié)將會(huì)引起的，所以，扭轉(zhuǎn)常數(shù)的橫截面，重力是研究裝系統(tǒng)來(lái)研究，采取扭轉(zhuǎn)剛度的垂直切片lxx不變的各個(gè)環(huán)節(jié)和梁作為設(shè)計(jì)變量的變化，從 0.1 x 105mm4 與 3.5 x 105 mm4。 圖-3 不斷的效果在垂直變形扭轉(zhuǎn) 圖-3顯示了平均位移與截面扭轉(zhuǎn)常數(shù)末端的各個(gè)環(huán)節(jié)和梁，根據(jù)它的變化速率的環(huán)節(jié)，是最大的，AB是鏈接，LJ依次分別GK梁和KM有在豎向剛度性能。其他的仿真結(jié)果表明，水平位移之間的差異進(jìn)行比較，結(jié)果表明該模型體育智力H和剛性模型變化小就改變了恒定不變的時(shí)候扭加載慣性力的線性驅(qū)動(dòng)器，但是水平位移的變化，這意味著在這種模擬豎向變形的生產(chǎn)水平位移系統(tǒng)機(jī)械手。注意端面線性驅(qū)動(dòng)器的主要原因是水平變形、線性驅(qū)動(dòng)器機(jī)器人是由兩個(gè)節(jié)點(diǎn)C和H . 所以，我們計(jì)算了不同的Z-coordinate攝氏度之間，如圖所示，在圖4 -扭轉(zhuǎn)常數(shù)的影響差別的鏈接德。其次是最有效的通用和連接梁，連接O3F，梁GK有效果。 因此，應(yīng)采取AB和連接區(qū)段大扭常數(shù)的免疫力，豎向剛度較大并行扭轉(zhuǎn)不變的鏈接德也使較少的均勻性，降低線性驅(qū)動(dòng)器不可以降低水平變形。 圖-4 在不影響扭不變 如圖-5、6所展示的影響是區(qū)域慣性轉(zhuǎn)矩的設(shè)計(jì)變量是區(qū)域剛度和慣性轉(zhuǎn)矩的各個(gè)環(huán)節(jié)和梁lz，圖顯示增加lw卡爾減少的速度高于垂直位移的不斷增加Ixx扭轉(zhuǎn)。這個(gè)Yxx AB、梁的鏈接，鏈接O3F是Iyy三個(gè)主要因素決定了豎向剛度。 圖-6 所示 鏈接的AB、梁公里，連接03F也是其中的三個(gè)主要因素決定的均勻性線性傳動(dòng)裝置、不同的分析結(jié)果表明，Izz效果好，具有至少兩個(gè)垂直和水平剛度，這意味著這種結(jié)構(gòu)，具有足夠的水平，降低Izz剛度的鏈接和增加Iyy AB、梁的鏈接，鏈接O3F公里的好方法，優(yōu)化系統(tǒng)。 圖-5 瞬間的慣性效應(yīng)對(duì)垂直位移 圖-6 轉(zhuǎn)動(dòng)慣量不平衡的影響 2.2影響的線性驅(qū)動(dòng)器的相對(duì)位置 線性執(zhí)行器的慣性是主要載荷之一，在機(jī)械手的運(yùn)動(dòng)，不同的相對(duì)應(yīng)的垂直位置產(chǎn)生不同的變形，圖7顯示了絕對(duì)平均的最終效應(yīng)垂直位移時(shí)驅(qū)動(dòng)馬達(dá)以恒定的加速度旋轉(zhuǎn)，我們可以看到，過(guò)低或過(guò)高的相對(duì)位置會(huì)造成比格變形，最好的位置是一對(duì)Z = 24毫米的地方大概是從中間環(huán)節(jié)連接O3F到 AB. 圖-7 影響線性驅(qū)動(dòng)器的相對(duì)位置 3 分析改進(jìn)的機(jī)械手 根據(jù)上述模擬結(jié)果，所有改進(jìn)的機(jī)械手的設(shè)計(jì)，時(shí)間如下：鏈接截面AB，DE，lJ 與30mm的基礎(chǔ)和高度，10毫米的厚度;鏈接O3F和矩形空心梁與30mm的基礎(chǔ)和高度工型鋼，l0mm法蘭和6mm網(wǎng);梁競(jìng)，通用汽車與8mm的堅(jiān)實(shí)基礎(chǔ)和30mm高的矩形。 圖-8 梯形運(yùn)動(dòng)姿態(tài) 圖-9中回應(yīng)的是機(jī)械手，相比之下，圖-10中提高初始的反應(yīng)，在其中所有的鏈接和機(jī)械手的矩形截面梁的堅(jiān)實(shí)基礎(chǔ)，用30毫米，高度的差異是曲線，C和H的曲線積分，二是垂直位移的末端，改進(jìn)系統(tǒng)中最大位移0.7Um最初的0.12Um相比，爭(zhēng)論的振動(dòng)激勵(lì)后仍停留在O.06Um±0.15% s±O.05Um相比的初始變形改善系統(tǒng)的初始小于前者具有較少的慣性，因?yàn)樵谙嗤牟椒ゲ粩嗉涌?，保持振?dòng)瓣膜差不多一樣，它對(duì)這整個(gè)系統(tǒng)中來(lái)說(shuō)，仍然改善系統(tǒng)的剛度，幾乎相當(dāng)于初始制度，針對(duì)大規(guī)模的平面并聯(lián)機(jī)構(gòu)在該系統(tǒng)相比下降了30%，這樣的初始優(yōu)化是有效的。 圖-9 、 圖-10 動(dòng)態(tài)響應(yīng) 4 結(jié)論 本文設(shè)計(jì)了一種新型三自由度機(jī)械手變量的敏感性進(jìn)行了研究在ADAMS環(huán)境中，可以得出以下結(jié)論： 1） 機(jī)器人具有較大的水平剛度，最終水平位移，效應(yīng)主要是由機(jī)械手垂直變形造成的，因此，更重要的是增加的幅度比剛度豎向剛度。 2） 參數(shù)Ixx，Iyy并鏈接'截面剛度Izz有不同的效應(yīng)，Iyy已經(jīng)對(duì)垂直剛度的影響最大，Ixx在第二位的是，Ixx具有在垂直剛度的影響最小，他們都較少對(duì)水平比垂直剛度剛度。 3） 橫截面的不同環(huán)節(jié)都有不同的影響，連線豎向剛度AB和德應(yīng)該使用區(qū)扭轉(zhuǎn)常數(shù)和慣性力矩大，如變形、長(zhǎng)方形、橫梁KM，，線 03F應(yīng)該使用區(qū)段形梁等重大時(shí)刻轉(zhuǎn)動(dòng)慣量、橫梁GK，和GM 可以使用盡可能的一小部分，從而降低了質(zhì)量。 4） 最佳的線性驅(qū)動(dòng)器的相對(duì)位置可以減少變形，最好的位置是垂直的平行結(jié)構(gòu)。 5） 改進(jìn)的機(jī)械手的動(dòng)態(tài)分析表明該優(yōu)化設(shè)計(jì)方法研究的基礎(chǔ)上的效率。 參考文獻(xiàn) [ l ] Dasgupta B，Mmthyunjayab T S。 The Stewart platform manipulator：a review。Mechat~m and Machine Theory，200o。35 (1)：15—40 [ 2 ] Xi F，Zhang D，Xu Z，et al。A comparative study on tripod u ts for machine Lo0ls。Intemational Journal of Machine TooLs&Manufacture，2003，43(7)：721—730 [ 3 ] Zhang D，Gosselin C M。Kinetostatic analysis and optimization of the Tricept machine tool family。In：Proceedings of Year 2000 Parallel Kinematic Machines International Conference，Ann Arbor，Michigan ，2001， 174—188 [ 4 ] Gosselin C M，Angeles J。A globe preference index for the kinematic optimum of robotic manipulator。ASME Journal of Mechanical ，l991，113(3)：220—226 [ 5 ] Gao F，I，iuX J，GruverW A。Performance evaluation of two-degree-of-freedom planar parallel robots。Mechanism and Machine Theory，l998，33(6)：661-668 [ 6 ] Huang T，Li M，Li Z X，et al。Optimal kinematic design of 2- DOF oaralel manipulator with well shahed workspace bounded by a specified conditioning index IEEE Transactions of Robot and Automation，2004，20，(3)：538—543 [ 7 ] Gosselin C M，Wang J。singularity loci ofplanarparallel manipulator with revoluted actuators。Robotics and Autonomousm，1997，2l(4)：377 398 [ 8 ] Yiu Y K，Cheng H，Xiong Z H，et al。on the dvnamies of Parallel Mmfipulators Proc。Of IEEE Inemational conference on Robotics& Automation。20o1。3766 3771 [ 9 ] Chakarov D。Study ofthe antagoni~ie stifness of parallel manipulators with actuation redundancy。Mechanism and Machine Theory，2004，39(6)：583—60l [ 10 ] Shab~a A A。Dynamics of Multibody systems。 Cambridge：Cambridge university press，l998。270-3 l0 11 維普資訊 http://www.cqv1p.com HIGH TECHNOLOGY LEITERS IVol. 12 No. 1 1 Jan. 2α)6 63 Dynamics optimization of a novel high speed and high precision 3-DOF manipulator① Lan Peng （蘭 朋悶，U Nianli , Sun Lining 鈴 ，Ding Qingyong 費(fèi) ( School of Me.chatroni 臼 Engineering, Har也m Institute of Te<丁hnology, Harbin 15α刷 ，P . R .China) ( ? Robotics Institute, H缸bin Institute of Technolo町，Harbin 15僅見(jiàn)I , P.R.China) Absti古董ct After introducing a novel 3-DOF high speed and high precision manipulator which combines direct driven planar parallel mechanism and linear actuator, ways of increasing its stiffness a陀 studied through dynamics simulation in ADAMS softw缸它 environment . Design study is carried out by parametric analysis tools to analyze the approximate sensitivity of the design V缸iables , including the effects of p缸沮netens of each beam cross section and relative position of linear actuator on model performance. Conclusions a陀 drdwn on the appropriate way of dynamics optimization to get a lightweight and small deformation manipu- lator. A planar parallel mechanism wi出 different cross section is used to an improved manipulator. Re- suits of dynamics simulation of the improved system and another unrefined one 缸e compa配d . η1e sti旺－ ness of them is almost equal , but the mass of 由e improved one decreas臼 greatly , which illustrates the ways efficient . Key words: manipulator, ADAMS, optimization , dynamics simulation 0 Introduction Parallel kinematic manipulatons ( PKM ） 耐 a class of promising machine for the manipulation and assemble of electronic device, because they have some advantages over the serial manipulator, such as high load ca町ing capacity , g0<對(duì) dynamic performance and precise position- ing[1l . A novel hybrid 3-DOF manipulator, which com- bines the advantages of parallel manipulator and serial manipulator, is studied in this paper. As shown in Fig. 1, the manipulator is composed of planar parallel mechanism ( PPM ) including parallelogram structure and linear actuator mounted on the end of PPM . Two A巳 di- 陀ct driven moto囚 integrated high 陀田lution em selected as driven part of planar par茍Ilel mechanism . Lln- 回 actuator is driven by voice coil motor, which is con- sidered as ideal driven part for short travel . As a kind of non-commutated di陀ct 世ive , hysteresis-free, cog-free device, voice coil motor can provide both high 歸sition sensitivity 阻d pert<叩force vensus stroke character. Hi班 precision linear encoder is used as feedback parts to 伊ar- ante悟 出e 陀陰暗tability in vertic況Idirection . Compared with higher degree of freedom parallel ma- nipulator, for example Steward platform or Tricept robot , kinematic and d嚴(yán)1amic m叫els of PPM ar吃 simple[ i-3] _ On the other hand , it has higher stiffness 由m 出e serial manipulator because of its close loop feature and low mo- ment of inertia . Meanwhile, the system can ove陀ome the mechanical elasticity introduced by flexible coupling, gear teeth, be白噸，bearing support , connecting shaft and other parts included by classical drive system . So this ma- nipulator is more easily to get well dynamics perfonnance and high p即ision . Planar parallel mechanism Motor Fig.1 3-00F hyhrid structUJ它 manipulator When the length of each link of pl四ar parallel mechanism is d配ided by kinematics analysis and syn出e- sis[4-7l , the primary task of optimal desi伊 should be in- creasing the stiffness and dee陀asing 由e mass. With re- gard to model wi由several par淚neters, it is important and 咀1at makes real time control possible and is mo陀 precise . neees ry to study the influence of each p neter on 田 缸田 ① Supported by 配 High Technology Research and Development Pr咱出nme of China ( No. 2003AA刷刷）） ． ② To whom coπ呵lOndence 動(dòng) uld be addressed . E而 mail: l皿 p@ sma. 凹陽(yáng) Ri,cpjved on Sept. 29，刻Xl4 維普資訊 h吐p://www .叫v1p.com 64 1-DGH TECHNOLO(,Y LF.ITERSI Vol.12 No. I I Jan. 2(脅 model performance in order tu make fu出er optimization . This paper will carry out design study by parametric anal- ysis tools of ADAMS, and then p陀sent appropriate ways of optimization to get a lightweight and small deformation system . 1 Simulation model ADAMS ( Automatic Dynamic Analy附 of Mechanical System) is a perfect software tool for dynamics simulation of mechanical system . It can deal with mechanism con- sisting of both rigid and flexible parts . Simulation model of the manipulator can be created in the ADAMS environ- ment as shown in Fig.2. OXYZ is the global reference frame , and o.切:yz is local reference frame. Two AC direct driven motors, expressed as 01A and Oi M , and linear actuator CH are t陀ated as rigid 以,dy . 幣1e rotor inertia of motor is 120kg · cm2 . 幣1e mass of linear actuator is 1.5kg. Links AB, DE, 03F and U are tr臼ted a,;; flexible effector is used to characterize the dynamic stiffness, which is different at different configuration during the lin- ear actuator moving fmward from initial position to the destination . 咀1e average vertical displacement of the end- effector is taken as the ob ective to study vertical stiff- n白白. The average difference of X-coordinate and Y『coor- dinate of point H between 由is model and a rigid model is 時(shí)cen as the objective to study level 叫iffness. 2 .1 Effect of er鴨 ”ction 。 Torsion defonnations of links will cause vertical dis- placement of the end-effector. So, torsional constants of cross section are studied first . Gravity is loaded on the system to study the vertical stiffness . Take torsional con- stant ι f section of each link and beam as design vari- able which varies from 0.1 x la5mm4 to 3 .5 x Ia5mm4 . Fig. 3 shows the average displacement of 由e end-effector body. Beam『 GK , GM and KM , which form a triangle, 即e also treated as flexible body . The length of links a陀 decided in advance by kinematics design a,;; AB ＝的F """ 7cm, DE = IJ = 7cm , GK = 7cm , GM = 11.66cm, KM = 8.338cm . 響1e other dimensions in the figure a陀 01 A = 02 M = 7cm, CB = CD = HJ ::: 2.5cm , EF = EC = JK = 3cm. Although the gross motion of planar parallel mecha- nism is in level, ho由 the vertical and level stiffness has to be considered . And the vertical stiffness is usually less than the level stiffness because of its cantilever character in vertical plane. 咀1e cross section size of each beam of -0.00相 r句Q士、 -0. 0013 ?！蕖? ，，-0.00’4 ?！匏? -0.00咀 -0.00’7 。由17 ?！尴? 0.0 0 $ ，．-- 壺 ’‘’ 1-link 03F 2-link AB 3-link DE ←→且也日 5-beam GK 6一戰(zhàn)am GM 7 斗:,eam KM ，。 ’5 2.0 2.5 3.0 3.6 T惆ional cons陽(yáng)ts (105mm勺 planar parallel mechanism and the relative position of lin- ear actuator 缸e two important factors that affect the stiff- ness of the system. Therefore, the following study is done in these respects . Fig.2 Simulation mudd 2 Simulation r四時(shí)t In this section , the average displacement of the end- Fig. 3 Effect of torsional constant on vertial deformation versus cross section torsional constant of each link and be卻n . According to it , the change rate of link AB is the biggest . Next is link DE ，日in tum res嚴(yán)ctively . B創(chuàng)ms GK and KM have the least eff白t on vertical stiffness. Other simulation shows that level displacement difference of point H between this model and a rigid model changes little with respect to a change in the torsional constant when constant level inertia force is loaded on the linear actuator. But the level displacement of the end-effector changes in 出is simulation . η1at means vertical deforτna- tion of the system should produce level displacement of end回effector. Note that unevenness of the linear actuator is the main cause of level defonnation of end-effect肘，and the linear actuator is supported by two joints C and H. So we calculate the difference of Z -coordinate between 陽(yáng)int C and H . As shown in Fig. 4, torsional constant of link DE affects the difference 出e most efficiently . Next is k田n GM and link U in order. Link 03F and beam GK have the leai;;t effect . 咀1erefore, links AB and DE should adopt se<;tion 維普資訊 h即／八il'WW.cqv1p .com HIGH TECHNOLC見(jiàn)Y IEITERSI Vol.12 No. l lJan. 2α)6 65 with big torsional constant to enhance the vertical stiff- ness. Bigger torsional constant of link DE also caus臼less unevenness of the linear actuator . Decreasing the uneven- ness can reduce the level deformation . 2-link AB  stant ι ．四e Irr of link AB , beam KM and link 03F are 出e three main factors that decide the vertical stiffness. Fig. 6 shows the Irr of link AB , beam KM and link 03F 缸它 also the three main facto陀 that decide the unevenness of the linear actuator. Di征erent analysis shows that I,, has the least effect on b 由 vertical and level stiffness . 幣iat means this kind of structure has enough level stiffness. So 4一UnmkUGM d asing I,, of links and increasing ”’ of link AB, －be晦 配陀 beam KM and link 03F 缸它 the good ways to optimize the system . 2.2 Effect of the relative position of linear ac陽(yáng)ator 3.7 3.6  QO 0.5 1.0 1.5 2.0 2.5 3.。 3.5 Torsional constants (1l?mm4) η1e inertia of linear actuator is one of the main loads during the motion of manipulator . Different relative ve民i- cal position should produce different deformation . Fig. 7 Fig.4 Effect of torsional constant on unevenness What Fig聲 .5 and 6 show are the effects of area m G （EE一守LD己 言ωERE昏時(shí)唱g－ 百多 .0.αJ05 shows the absolute average vertical displacement of end- effector when the driven motors rotate at a constant accel- eration . We can see that too low or too hi班 時(shí)ative posi- tion will cause bigger defonnation ．咀1e best position is at a如ut Z = - 24mm where is approximate the midst from link AB to link 03F. m E 。 ’6 m A .0.0025。。 4一link u 6 beam GM 0.5 1.0 1.5 Moments of iner由（llfmm‘｝  J d O D41 l AB 亞 刷 E u 2.0 2.5 ] Fig.5 Effect of moments of i陽(yáng)rtia on vertial defo回國(guó)lion 。。  0.0 -?J.0  -40.0 .10.0 Z-coordinate (mm)  篇。 eoo 1 」ink 03F 3---link DE 5一悅am GK 2一link:AB 4--link 日 6--bE姐m GM  Fig.7 Effect of relative position of linear actuator 3 --5 ＂’ ι乙 』 回南.：：：.－..乓芫叫 .. · 、、 3 Analysis of an improved manipulator According to above simulation result , an improved manipulator is designed as follows : cross s創(chuàng)ions of links AB , DE ，日在附 hollow rectangle with 30mm b出e and 2.5 0.0 0.5 1.0 1.5 2.0 2.5 height , 10mm thickne制；link 向F and beam KM 町e I- Moments of inertia (105mm‘） Fig. 6 Effect of moments of inertia on unevenness moment of inertia on the stiffness. 'lbe design variable is 臟a moment of inertia lyy and ι of each link and beam . Fig. 5 shows that inc陀兇e of Irr can 配duce the vertical deformation more rapidly than increase of torsional con- beam with 30mm ba

收藏