摘 要

　　近幾年,軟體機器人技術(shù)迅猛發(fā)展,其以軟體生物為參考,基于智能軟材料研發(fā)高自由度的軟體機器人.相比傳統剛性機器人,這些機器人具有環(huán)境適應能力強、控制簡(jiǎn)單等優(yōu)勢,從而使其在工業(yè)、醫療和軍事等領(lǐng)域發(fā)展潛力巨大.軟體驅動(dòng)器技術(shù)是軟體機器人的核心技術(shù)之一.在眾多軟體驅動(dòng)器中,氣動(dòng)軟體驅動(dòng)器柔順性好、功率質(zhì)量比高、制造成本低等諸多優(yōu)點(diǎn)使其成為研究的熱點(diǎn).本文以氣動(dòng)軟體驅動(dòng)器為研究對象,對此類(lèi)驅動(dòng)器進(jìn)行設計與制造,并基于該類(lèi)驅動(dòng)器研制三款軟體操作手.

　　氣動(dòng)軟體驅動(dòng)器的非線(xiàn)性運動(dòng)學(xué)響應是其設計的難點(diǎn),采用有限元方法對纖維增強型和多腔體型兩種典型的氣動(dòng)軟體彎曲驅動(dòng)器的驅動(dòng)性能與結構參數之間的關(guān)系進(jìn)行仿真分析.結合仿真結果,采用 3D模型法制作兩類(lèi)彎曲驅動(dòng)器.為實(shí)現多驅動(dòng)器協(xié)作驅動(dòng),設計軟體驅動(dòng)器的控制系統,該控制系統獨立控制多驅動(dòng)器系統中每個(gè)驅動(dòng)器完成彎曲、保持以及伸展動(dòng)作.將以上氣動(dòng)軟體驅動(dòng)器的研究成果應用于三款軟體操作手,即:工業(yè)軟體抓手、軟體仿人手和軟體康復手套.軟體工業(yè)抓手抓取穩定、自適應能力強,配合 UR 機械臂能夠實(shí)現多種物品抓取.結合數據手套或肌電手環(huán),實(shí)現軟體仿人手靈巧動(dòng)作復現.設計具有敞開(kāi)式結構的軟體康復手套,輔助實(shí)現手功能障礙患者的主動(dòng)康復訓練.最后,開(kāi)發(fā)可視化界面為軟體仿人手展示和軟體康復手套使用提供便利.

　　本文通過(guò)對新興軟體機器人技術(shù)的研究,對典型氣動(dòng)軟體彎曲驅動(dòng)器的驅動(dòng)性能與其截面尺寸等參數的關(guān)系有了更深入的了解,開(kāi)發(fā)了三款基于氣動(dòng)軟體彎曲驅動(dòng)器軟體操作手系統.這些研究成果對軟體機器人技術(shù)今后的應用及進(jìn)一步發(fā)展具有積極的作用.

　　關(guān)鍵詞:軟體機器人、氣動(dòng)軟體驅動(dòng)器、驅動(dòng)器控制、軟體操作手

　　ABSTRACT

　　In recent years, soft robotics based on soft smart material hasdeveloped rapidly. Bioinspired by living organism, many robots with highdegree of freedoms have been developed. Compared with traditional rigidrobots, these soft robots can inherently provide adaptable morphology inresponse to environmental changes and produce sophisticated motions withsimple controls, which makes them have great potential value inautomation, medicine and military domains. One of the core techniques ofsoft robotics is the soft actuator technique. Pneumatic actuators are ofparticular interest because they have many advantages, such as highcompliance, high power to weight ratio and easy fabrication with emergingdigital fabrication techniques. This thesis focuses on the design andfabrication of pneumatic actuators. Using these actuators, three types ofsoft manipulators have been developed.

　　The significant potential of pneumatic soft actuators is currentlylimited by their nonlinear kinematic responses. In this paper, finite elementanalysis was used as a design tool to find the optimal geometric parametersfor two typical pneumatic soft bending actuators which have receivedsignificant research attention recently. Combined with the simulationresults, molding method is used to fabricate actuators. The control systemhas been accomplished for achieving multiple actuators control. Everysingle actuator can realize bending, maintaining and extending with thehelp of the designed control system. Three soft manipulators including softautomate gripper, soft hand and soft rehabilitation glove were developedbased on the research results. The soft automate gripper has a strongadaptive ability. Cooperated with UR robot arm, it can grasp differentABSTRACTkinds of objects. Combined with data glove or EMG band, the soft handcan display dexterous actions similar as the human hand. The open-palmdesigned soft rehabilitation glove can help hand functional disorderedpatients do active training. At last, the visual interface was developed tofacilitate the display of soft hand and the application of soft rehabilitationglove.

　　According to this research, a better understanding of the relationshipbetween the typical pneumatic soft bending actuators and their structureparameters has been obtained and three soft manipulator systems based onthese bending actuators have been developed. These research results willhave a positive effect on the application and further development of softrobotics in the future.

　　KEY WORDS: soft robotics, pneumatic soft actuator, control of softactuators,soft manipulator

　　目 錄

　　摘 要············································································I

　　ABSTRACT··································································III

　　第一章 緒論···································································1

　　1.1 課題來(lái)源 ··································································1

　　1.2 研究背景與意義 ···························································1

　　1.3 軟體驅動(dòng)器研究現狀 ·····················································2

　　1.3.1 電響應軟體驅動(dòng)器 ·····················································3

　　1.3.2 磁響應軟體驅動(dòng)器 ·····················································3

　　1.3.3 化學(xué)響應軟體驅動(dòng)器 ··················································4

　　1.3.4 熱響應軟體驅動(dòng)器 ·····················································5

　　1.3.5 光敏軟體驅動(dòng)器 ························································5

　　1.3.6 氣動(dòng)軟體驅動(dòng)器 ························································6

　　1.4 論文內容及章節安排 ····················································7

　　第二章 氣動(dòng)軟體驅動(dòng)器工作原理 ·········································8

　　2.1 驅動(dòng)原理 ···································································8

　　2.1.1 纖維增強型氣動(dòng)軟體驅動(dòng)器 ········································8

　　2.1.2 多腔體型氣動(dòng)軟體驅動(dòng)器 ···········································11

　　2.2 理論分析 ··································································11

　　2.2.1 纖維增強型驅動(dòng)器的理論分析 ····································11

　　2.2.2 多腔體型驅動(dòng)器的理論分析 ·······································18

　　2.3 有限元仿真分析 ························································19

　　2.3.1 纖維增強型驅動(dòng)器的有限元分析 ·································19

　　2.3.2 多腔體型驅動(dòng)器的有限元分析 ····································23

　　2.4 本章小結 ·································································29

　　第三章 氣動(dòng)軟體驅動(dòng)器制備 ·············································30

　　3.1 制作工藝 ································································30

　　3.2 纖維增強型驅動(dòng)器 ······················································31

　　3.2.1 實(shí)驗材料與儀器 ·····················································32

　　3.2.2 制作過(guò)程······························································32

　　3.2.3 性能測試······························································34

　　3.3 多腔體型驅動(dòng)器························································36

　　3.3.1 實(shí)驗材料與儀器······················································36

　　3.3.2 制作過(guò)程······························································36

　　3.3.3 性能測試······························································38

　　3.4 本章小結·································································39

　　第四章 驅動(dòng)器控制系統設計···············································40

　　4.1 方案設計··································································40

　　4.2 硬件系統搭建····························································40

　　4.2.1 硬件選型·······························································42

　　4.2.2 系統搭建·······························································44

　　4.2.3 通訊協(xié)議······························································44

　　4.3 軟件系統開(kāi)發(fā)···························································46

　　4.3.1 自定義指令格式······················································46

　　4.3.2 控制邏輯·······························································46

　　4.3.3 人機交互界面·························································48

　　4.4 本章小結·································································49

　　第五章 軟體操作手集成與應用·········································50

　　5.1 應用簡(jiǎn)介·································································50

　　5.2 軟體工業(yè)抓手···························································50

　　5.2.1 抓手的設計···························································50

　　5.2.2 抓取功能展示·························································53

　　5.3 軟體仿人手······························································54

　　5.3.1 組成結構·······························································54

　　5.3.2 計數手勢展示··························································56

　　5.3.3 結合數據手套的展示················································57

　　5.3.4 結合肌電手環(huán)的展示···············································59

　　5.4 軟體康復手套···························································61

　　5.4.1 手套的組成結構·······················································61

　　5.4.2 軟體康復手套的應用················································62

　　5.5 本章小結·································································64

　　第六章 總結與展望 ··························································65

　　6.1 主要工作與創(chuàng )新點(diǎn) ······················································65

　　6.2 后續研究工作 ···························································65

　　參 考 文 獻····································································66

　　附錄 1 ··········································································70

　　附錄 2 ········································································73

　　致 謝 ·········································································75

　　攻讀碩士學(xué)位期間已發(fā)表或錄用的論文·······························76

　　第一章 緒論

　　1.1 課題來(lái)源

　　本課題研究?jì)热輥?lái)源于國家自然科學(xué)基金優(yōu)秀青年科學(xué)基金項目"軟體機器人設計與控制"(項目編號:51622506)以及上海市"科技創(chuàng )新行動(dòng)計劃"基礎研究領(lǐng)域項目"基于軟體智能材料的類(lèi)人靈巧手設計與控制"(項目編號:16JC1401000).

　　1.2 研究背景與意義

　　機器人學(xué)是一門(mén)研究如何構建滿(mǎn)足所需運動(dòng)、感知等能力機器人的科學(xué),如 今的機器人已經(jīng)成為科學(xué)研究的重要工具.以工業(yè)機器人為代表的傳統機器人在運動(dòng)的復雜性、控制的準確性等方面已經(jīng)取得了前所未有的進(jìn)展.但是,人們依 然期待著(zhù)適用于更多應用場(chǎng)景以及可以完成更加復雜任務(wù)的新型機器人的出現.

　　在現階段機器人研究中,如何將"軟"應用于構建新型機器人是推動(dòng)機器人進(jìn)一步發(fā)展所面臨的挑戰.解決好此類(lèi)問(wèn)題將會(huì )大大加快新型機器人的研發(fā)進(jìn)程,進(jìn)而誕生出應用領(lǐng)域更廣泛的新一代機器人 [1].

　　具有彈性或柔順性軟體結構的機器人與外界接觸時(shí)能夠自動(dòng)做出變形等動(dòng)作,可以更好地應用仿生學(xué)研究成果.因此,越來(lái)越多的軟體結構在機器人系統的設計中被采用.自然界中的生物利用自身柔軟的組織和適應性的結構可以在復雜的自然界中高效地活動(dòng),所以對這些生物的研究可以揭示出很多有利于"軟體機器人"拓展能力以及提高任務(wù)執行效率很有意義的準則."軟體機器人"一詞曾被用來(lái)表示具有剛性連桿和機械(或被動(dòng))柔性關(guān)節、具有可變剛度、采用柔順或阻抗控制的機器人[2-3];之后,該詞被用于強調從機器人"具有剛性連接"到"仿生連續體"的轉變[4]; 然后,Trivedi[5]等人認為軟體機器人不同于具有剛性連接的傳統機器人,也不同于具有多剛性關(guān)節的超冗余度蛇形機器人,"軟體機器人"一詞應該用來(lái)描述采用軟材料和柔性結構的新一代機器人.軟體機器人指依賴(lài)自身材料的彈性或結構的順應性可以經(jīng)歷大變形從而實(shí)現主動(dòng)地與環(huán)境進(jìn)行交互的機器人.其定義更加注重軟體機器人與環(huán)境交互時(shí)表現出的柔順性和變形能力.軟體機器人的柔順性使其可以自適應物體的復雜表面,接觸時(shí)吸收碰撞能保持系統穩定性,表現出物理上的魯棒性和人機交互的安全性,同時(shí)具有成本低廉的潛在優(yōu)勢.

　　圖 1-1 展示了軟體機器人技術(shù)發(fā)展圖譜,從純剛性機器人逐步發(fā)展到純軟體機器人.該圖譜羅列了多款主要采用軟材料和變形結構制作的機器人,它們的結 構中幾乎不含剛性元件,這種類(lèi)型的軟體機器人占據了今天軟體機器人中的大多數.

　　軟體機器人技術(shù)橫跨工程學(xué)、材料學(xué)、生物學(xué)、數學(xué)、醫學(xué)等諸多學(xué)科,多學(xué)科交叉的特性使其可以采用一些非常規的科學(xué)手段實(shí)現所需的功能.現已出現了擁有多種多樣能力的軟體機器人,它們不僅僅可以實(shí)現抓握、移動(dòng),而且可以做出諸如擠壓[6]、跳躍[7]、攀爬[8]以及生長(cháng)[9]等基于剛性連接傳統機器人無(wú)法實(shí)現的功能,這些獨特的功能為軟體機器人開(kāi)辟了全新的應用領(lǐng)域.

　　本論文主要研究氣動(dòng)軟體驅動(dòng)器的設計、制造、控制技術(shù)及其在軟體操作手上的應用.采用有限元方法對氣動(dòng)軟體驅動(dòng)器進(jìn)行仿真分析,設計并制造了纏線(xiàn)增強型和多腔體型兩類(lèi)氣動(dòng)軟體驅動(dòng)器;通過(guò)軟硬件系統開(kāi)發(fā),設計了氣動(dòng)軟體驅動(dòng)器的控制平臺;在此基礎上,試制了軟體工業(yè)抓手、仿人軟體手、軟體康復手套三個(gè)原型樣機系統,開(kāi)展了相應系統集成和實(shí)驗驗證工作,取得了較好的實(shí)驗效果.

　　1.3 軟體驅動(dòng)器研究現狀

　　根據軟體驅動(dòng)器激勵信號的不同,可以將其劃分為:電響應軟體驅動(dòng)器 [10-13]、磁響應軟體驅動(dòng)器 [14-21]、化學(xué)響應軟體驅動(dòng)器 [22] 、熱響應軟體驅動(dòng)器 [23、24] 、 光響應軟體驅動(dòng)器 [25-27] 以及壓力驅動(dòng)軟體驅動(dòng)器 [28-31].以下對這六種軟體驅動(dòng)器分別進(jìn)行介紹.

　　1.3.1 電響應軟體驅動(dòng)器

　　電響應軟體驅動(dòng)器主要涉及電致變形智能軟材料的應用,這些材料包括聚合物、凝膠、流體、紙張以及獨立的碳納米管等,它們可以實(shí)現電能到機械能的轉化.此類(lèi)軟體驅動(dòng)器可以通過(guò)改變電激勵信號幅度、相位以及頻率的方式來(lái)調制.此外,因為電響應軟體驅動(dòng)器與傳統的電子元器件等兼容,所以基于電子元器件的集成很容易實(shí)現.此類(lèi)驅動(dòng)器的研究重點(diǎn)有人工肌肉[11],微尺度對象操作[12]和微流控系統[13].

　　電響應軟體驅動(dòng)器的典型代表是介電彈性體驅動(dòng)器,驅動(dòng)器由兩側覆蓋柔性電極的介電彈性體薄膜組成,施加驅動(dòng)電壓時(shí),介電彈性體薄膜在電場(chǎng)力作用下產(chǎn)生變形.雖然采用這種方式設計制作的軟體驅動(dòng)器具有大的驅動(dòng)力和驅動(dòng)位移, 但其驅動(dòng)電壓往往要數千伏,這是阻礙其實(shí)際應用的最大挑戰.

　　1.3.2 磁響應軟體驅動(dòng)器

　　磁響應軟體驅動(dòng)器主要涉及具有磁性的智能軟材料.典型地,將磁性顆粒或離散的磁體摻入軟體化合物中就可以制作出一款具有可變磁特性曲線(xiàn)的復合材料.當這種復合材料處于一個(gè)磁場(chǎng)中時(shí),內嵌的磁性顆粒或附帶的磁體就會(huì )試圖與該磁場(chǎng)對齊,從而產(chǎn)生力矩和變形.磁場(chǎng)強度和磁場(chǎng)的空間梯度可以在很小的空間獨立產(chǎn)生,因此可以認為磁場(chǎng)與磁顆粒或磁體相互作用時(shí)產(chǎn)生的驅動(dòng)力和驅動(dòng)力矩是解耦的.磁驅動(dòng)可以有效地產(chǎn)生驅動(dòng)力和驅動(dòng)力矩兩種獨立的驅動(dòng)作用進(jìn)而組合實(shí)現更復雜的驅動(dòng).此類(lèi)軟體驅動(dòng)器的變形模式可以通過(guò)改變驅動(dòng)信號、驅動(dòng)器整體形狀以及復合材料的磁特性曲線(xiàn)或剛度來(lái)實(shí)現不同的設計.

　　因為磁場(chǎng)可以穿透大部分材料,所以磁響應軟體驅動(dòng)器可用于封閉狹小空間,特別適用于像體內靶向投送藥物、顯微外科手術(shù)、微流控以及體內組裝等應用場(chǎng)景[15-17].相對其他驅動(dòng)形式,磁響應驅動(dòng)反應較快,頻率可以達到 100Hz[18],而 且其控制與裝配過(guò)程密切相關(guān)[19].現在,磁響應軟體驅動(dòng)器已經(jīng)成功地被應用于制作游泳機器人[14]、爬行器[20]和微型泵[21]等.

　　然而,由于磁力縮放的不合理等原因,外部驅動(dòng)的磁線(xiàn)圈往往比較大而且能耗很大.因此磁響應軟體驅動(dòng)器適用于所施加磁場(chǎng)和梯度具有顯著(zhù)強度并且易控的狹小工作空間.

　　1.3.3 化學(xué)響應軟體驅動(dòng)器

　　化學(xué)響應可以包含很多種響應機制,是一個(gè)相當廣泛的范疇.此處限定此類(lèi)驅動(dòng)器主要是以液體或蒸汽形式的激勵信號產(chǎn)生驅動(dòng),過(guò)程中常常涉及化學(xué)反應、誘導應力、誘導變形等.此外,還包括基于毛細管力的驅動(dòng)器,其作用于以水為代表的流體的引入和蒸發(fā).

　　化學(xué)能到機械能的轉化過(guò)程稱(chēng)為化學(xué)機械運動(dòng).像聚合物網(wǎng)絡(luò )、凝膠這樣的軟材料具有選擇性地定向擴散化學(xué)物質(zhì)的能力.材料中物質(zhì)的擴散會(huì )引起機械應力,同時(shí),物質(zhì)在材料內部擴散過(guò)程中會(huì )發(fā)生各種化學(xué)反應.由酸、堿、有機溶劑等刺激引起的反應可以改變聚合物鏈之間的相互作同等,這些變化可以改變滲透壓或增加鏈的親和性等,最終導致材料尺寸和形狀的變化.Hore 等人在一個(gè)機器人案例中應用這種原理展示了有意思的彈性滾筒上坡現象[22].其原理為: 當有機溶劑被添加到滾筒表面時(shí),它積聚在氣缸的下側,使其膨脹,從而以足夠的力將滾筒推向上,此推動(dòng)力可以承載滾筒自身重量的 8 到 10 倍.

　　一般來(lái)說(shuō),化學(xué)響應的軟體驅動(dòng)器往往是高度敏感的,其響應時(shí)間可變的,主要取決于擴散速率等.但此類(lèi)軟體機器人的應用需要特定的液體或潮濕的化學(xué)反應條件,同時(shí)還面臨著(zhù)驅動(dòng)力小和難以精確控制的挑戰.

　　1.3.4 熱響應軟體驅動(dòng)器

　　適用于熱響應軟體驅動(dòng)器的熱激勵信號主要包括(近)紅外光、熱輻射以及焦耳加熱.焦耳加熱就是靠導電材料通電做功產(chǎn)生熱.熱響應驅動(dòng)器可以使用激光等遠程施加熱激勵信號,小范圍或大范圍都可應用.只要溫度保持在 4-37℃之間此類(lèi)驅動(dòng)器就可以應用于活細胞[24],所以與溶劑和紫外線(xiàn)刺激相比熱激勵方式更安全.

　　然而,熱響應的驅動(dòng)器通常響應慢、效率很低.使用更薄的膜、更好的吸熱材料以及高功率在一定程度上可以提高其效率.

　　1.3.5 光敏軟體驅動(dòng)器

　　光敏軟體驅動(dòng)器可以遠程精確控制、快速調制,其應用主要集中在納米尺度和微尺度[26,27].光致變色分子在合成的光學(xué)系統中發(fā)揮著(zhù)重要作用,它們可以捕捉目標光信號并轉化為應變等驅動(dòng)信號.這種原理與自然界中的光驅動(dòng)機制類(lèi)似,如光誘導的視網(wǎng)膜分子異構化觸發(fā)一系列化學(xué)反應,最終產(chǎn)生神經(jīng)信號以及實(shí)現對光的感知.這些感光分子已經(jīng)在由聚合物、凝膠、流體以及光電材料等組成的微型軟體驅動(dòng)器中有了應用.

　　此類(lèi)驅動(dòng)器的應用主要受到可用的感光材料種類(lèi)少以及響應速度較慢等的限 制.

　　1.3.6 氣動(dòng)軟體驅動(dòng)器

　　氣動(dòng)軟體驅動(dòng)器根據驅動(dòng)方式的不同可歸類(lèi)為:伸長(cháng)驅動(dòng)器、收縮驅動(dòng)器、扭轉驅動(dòng)器和彎曲驅動(dòng)器[29].所有的這些驅動(dòng)都對驅動(dòng)器結構中剛度的空間分布進(jìn)行設計,從而使它們可以在氣壓下產(chǎn)生期望的變形.驅動(dòng)器材料一般選為硅膠.基于氣動(dòng)軟體驅動(dòng)器,哈佛大學(xué) George M. Whitesides 研究組研發(fā)出了充氣式蠕動(dòng)軟體機器人[28],日本岡山大學(xué)研發(fā)了氣動(dòng)軟體機器魚(yú)[30],北京航空航天大學(xué)機器人研究所研發(fā)了一款氣動(dòng)仿生蝠鲼機器魚(yú)[31].

　　研究表明,小型氣動(dòng)軟體驅動(dòng)器就可以產(chǎn)生較大的力.相比其它驅動(dòng)方式,這類(lèi)采用壓力信號激勵的驅動(dòng)器的應用所受限制因素很少.因此,以氣動(dòng)彈性體驅動(dòng)器為代表的氣動(dòng)軟體驅動(dòng)器在近幾年成為研究人員關(guān)注的焦點(diǎn).此類(lèi)驅動(dòng)器的應用不需要高的電、磁場(chǎng);主要材料是硅膠,容易獲得且成本低,并且對人體安全.此類(lèi)驅動(dòng)器的優(yōu)點(diǎn)可以概括為結構簡(jiǎn)單、變形能力強、柔順性好、功率質(zhì)量比高、響應速度快、制造成本低以及控制簡(jiǎn)單.所以,本研究選取氣動(dòng)軟體驅動(dòng)器作為研究對象.

　　1.4 論文內容及章節安排

　　本論文主要研究氣動(dòng)軟體驅動(dòng)器的設計、制造、控制技術(shù)及其在軟體操作手上的應用.采用有限元分析法對氣動(dòng)軟體驅動(dòng)器進(jìn)行仿真分析,設計并制造了纏線(xiàn)增強型和多腔體型兩類(lèi)氣動(dòng)軟體驅動(dòng)器;通過(guò)軟硬件系統開(kāi)發(fā),設計了氣動(dòng)軟體驅動(dòng)器的控制平臺;在此基礎上,試制了軟體工業(yè)抓手、軟體仿人手、軟體康復手套三個(gè)原型樣機系統.

　　論文的章節安排如下:

　　第一章,對軟體機器人技術(shù)進(jìn)行了簡(jiǎn)單介紹,基于不同類(lèi)型軟體驅動(dòng)器特點(diǎn)的分析,選取氣動(dòng)軟體驅動(dòng)器作為本論文的研究對象;

　　第二章,簡(jiǎn)單介紹了纖維增強型和多腔體型兩類(lèi)常見(jiàn)的氣動(dòng)軟體驅動(dòng)器的工作原理并進(jìn)行了理論分析,采用有限元方法對兩類(lèi)驅動(dòng)器的重要結構參數進(jìn)行分析;

　　第三章,概述了現有的氣動(dòng)軟體驅動(dòng)器制作工藝,選用 3D 模型法制作兩類(lèi)氣動(dòng)軟體驅動(dòng)器并進(jìn)行了相關(guān)的性能測試;

　　第四章,根據氣動(dòng)軟體驅動(dòng)器的工作特點(diǎn)設計了通用性的控制系統,主要包括硬件系統的搭建和軟件系統的開(kāi)發(fā);

　　第五章,基于以上研究成果試制了軟體工業(yè)抓手、軟體仿人手以及軟體康復手套原型樣機系統;

　　在第六章中,總結本研究中所取得的研究成果并規劃下一步研究可以展開(kāi)的方向.

　　…………由于本文篇幅較長(cháng),部分內容省略,詳細全文見(jiàn)文末附件

　　第六章 總結與展望

　　6.1 主要工作與創(chuàng )新點(diǎn)

　　本文對軟體機器人技術(shù)中的氣動(dòng)軟體驅動(dòng)器進(jìn)行了較深入的研究,主要包括對其工作原理的分析、驅動(dòng)器性能與其結構參數關(guān)系的有限元仿真分析、軟體驅動(dòng)器及其控制系統的設計制作以及基于氣動(dòng)軟體驅動(dòng)器的軟體工業(yè)抓手、軟體仿人手和軟體康復手套實(shí)現.

　　本文的創(chuàng )新點(diǎn)是在充分分析氣動(dòng)軟體驅動(dòng)器工作原理和特性的基礎上將其進(jìn)行應用,充分利用驅動(dòng)器自身的非線(xiàn)性?xún)?yōu)勢;設計出一套針對氣動(dòng)軟體驅動(dòng)器的通用性便攜式控制設備,并開(kāi)發(fā)了配套的人機交互軟件;該控制設備采用系統中集成微型電動(dòng)氣泵的方式解決了氣動(dòng)軟體系統因需要外接氣源而造成的系統不獨立問(wèn)題;制作的三款氣動(dòng)軟體操作手為對應的應用場(chǎng)景提供了不同于傳統機器人的全新解決方案 .

　　綜上,本研究對軟體機器人今后的應用及進(jìn)一步發(fā)展具有積極的推動(dòng)作用.

　　6.2 后續研究工作

　　盡管軟體機器人憑借有別于傳統剛性機器人的獨特優(yōu)勢得到了廣泛研究并取得了快速發(fā)展,但現階段純軟的軟體機器人還存在很多缺點(diǎn).軟體機器人技術(shù)依然可以通過(guò)向自然界的生物學(xué)習來(lái)進(jìn)行改進(jìn),如學(xué)習生物的形變能力以及生長(cháng)過(guò)程中展現出的對環(huán)境的適應性改變甚至其身體的自修復能力等.

　　如果仔細觀(guān)察自然中的生物體可以發(fā)現,純軟體的動(dòng)物往往比較小,而體型較大的動(dòng)物通常需要骨架結構來(lái)支撐整個(gè)身體的重量.雖然在水中(例如水母)或地下(如巨蚯蚓)也存在體型較大的無(wú)骨骼生物,但它們的身體往往被周?chē)�介質(zhì)所支撐.這一生物學(xué)證據表明在軟體機器人的研究中也可以嘗試將軟材料與剛性結構相結合這一設計理念.這種設計將會(huì )增加軟體機器人對環(huán)境施加的力等,可以大大增加它們的應用潛力.所以,接下來(lái)將嘗試將軟體機器人的設計中結合剛性結構來(lái)進(jìn)一步研究.

　　同時(shí),隨著(zhù)計算機技術(shù)的快速發(fā)展,接下來(lái)可以嘗試研究基于學(xué)習的控制方法來(lái)實(shí)現對具有復雜非線(xiàn)性行為軟體驅動(dòng)器的控制等.
　　參 考 文 獻
　　[1] Laschi C, Mazzolai B, Cianchetti M. Soft robotics: Technologies and systemspushing the boundaries of robot abilities[J]. Sci. Robot., 2016, 1(1): eaah3690.
　　[2] Albu-Schaffer A, Fischer M, Schreiber G, et al. Soft robotics: what cartesianstiffness can obtain with passively compliant, uncoupled joints[C]//IntelligentRobots and Systems, 2004.(IROS 2004). Proceedings. 2004 IEEE/RSJInternational Conference on. IEEE, 2004, 4: 3295-3301.
　　[3] Albu-Schaffer A, Eiberger O, Grebenstein M, et al. Soft robotics[J]. IEEE Robotics& Automation Magazine, 2008, 15(3).
　　[4] D. Trivedi, C. D. Rahn, W. M. Kier, I. D. Walker, Soft robotics: Biologicalinspiration, state of the art, and future research. Appl. Bionics Biomech. 5, 99-117(2008).
　　[5] Trivedi D, Lotfi A, Rahn C D. Geometrically exact models for soft roboticmanipulators[J]. IEEE Transactions on Robotics, 2008, 24(4): 773-780.
　　[6] Jayaram K, Full R J. Cockroaches traverse crevices, crawl rapidly in confinedspaces, and inspire a soft, legged robot[J]. Proceedings of the National Academyof Sciences, 2016, 113(8): E950-E957.
　　[7] Lin H T, Leisk G G, Trimmer B. GoQBot: a caterpillar-inspired soft-bodied rollingrobot[J]. Bioinspiration & biomimetics, 2011, 6(2): 026007.
　　[8] Kim S, Spenko M, Trujillo S, et al. Smooth vertical surface climbing withdirectional adhesion[J]. IEEE Transactions on robotics, 2008, 24(1): 65-74.
　　[9] Sadeghi A, Tonazzini A, Popova L, et al. A novel growing device inspired by plantroot soil penetration behaviors[J]. PloS one, 2014, 9(2): e90139.
　　[10] Christianson C, Goldberg N N, Deheyn D D, et al. Translucent soft robots drivenby frameless fluid electrode dielectric elastomer actuators[J]. Science Robotics,2018, 3(17): eaat1893.
　　[11] Hines L, Petersen K, Lum G Z, et al. Soft Actuators for Small‐Scale Robotics[J].
　　Advanced Materials, 2017, 29(13).
　　[12] Jager E W H, Ingan?s O, Lundstr?m I. Microrobots for micrometer-size objects inaqueous media: potential tools for single-cell manipulation[J]. Science, 2000,288(5475): 2335-2338.
　　[13] Raghavan R V, Qin J, Yeo L Y, et al. Electrokinetic actuation of low conductivitydielectric liquids[J]. Sensors and Actuators B: Chemical, 2009, 140(1): 287-294.
　　[14] Diller E, Zhuang J, Zhan Lum G, et al. Continuously distributed magnetizationprofile for millimeter-scale elastomeric undulatory swimming[J]. Applied PhysicsLetters, 2014, 104(17): 174101.
　　[15] Diller E, Giltinan J, Lum G Z, et al. Six-degree-of-freedom magnetic actuation forwireless microrobotics[J]. The International Journal of Robotics Research, 2016,35(1-3): 114-128.
　　[16] Kummer M P, Abbott J J, Kratochvil B E, et al. OctoMag: An electromagneticsystem for 5-DOF wireless micromanipulation[J]. IEEE Transactions on Robotics,2010, 26(6): 1006-1017.
　　[17] Tasoglu S, Diller E, Guven S, et al. Untethered micro-robotic coding ofthree-dimensional material composition[J]. Nature communications, 2014, 5:
　　3124.
　　[18] Garstecki P, Tierno P, Weibel D B, et al. Propulsion of flexible polymer structuresin a rotating magnetic field[J]. Journal of Physics: Condensed Matter, 2009,21(20): 204110.
　　[19] Diller E, Giltinan J, Sitti M. Independent control of multiple magnetic microrobotsin three dimensions[J]. The International Journal of Robotics Research, 2013,32(5): 614-631.
　　[20] Khoo M, Liu C. Micro magnetic silicone elastomer membrane actuator[J]. Sensorsand Actuators A: Physical, 2001, 89(3): 259-266.
　　[21] Lee H, Kim J, Kim J, et al. Programming magnetic anisotropy in polymericmicroactuators[J]. Nature materials, 2011, 10(10): 747.
　　[22] Hore D, Majumder A, Mondal S, et al. How to make a cylinder roll uphill[J]. SoftMatter, 2012, 8(18): 5038-5042.
　　[23] Taccola S, Greco F, Sinibaldi E, et al. Toward a new generation of electricallycontrollable hygromorphic soft actuators[J]. Advanced Materials, 2015, 27(10):
　　1668-1675.
　　[24] Stroganov V, Zakharchenko S, Sperling E, et al. Biodegradable Self‐FoldingPolymer Films with Controlled Thermo‐Triggered Folding[J]. AdvancedFunctional Materials, 2014, 24(27): 4357-4363.
　　[25] Yamada, Munenori, et al. "Photomobile polymer materials: towards light‐drivenplastic motors." Angewandte Chemie International Edition 47.27 (2008):
　　4986-4988.
　　[26] Ercole F, Davis T P, Evans R A. Photo-responsive systems and biomaterials:
　　photochromic polymers, light-triggered self-assembly, surface modification,fluorescence modulation and beyond[J]. Polymer Chemistry, 2010, 1(1): 37-54.
　　[27] Wei J, Yu Y. Photodeformable polymer gels and crosslinked liquid-crystallinepolymers[J]. Soft Matter, 2012, 8(31): 8050-8059.
　　[28] Shepherd, Robert F., et al. "Multigait soft robot." Proceedings of the nationalacademy of sciences 108.51 (2011): 20400-20403.
　　[29] Gorissen B, Chishiro T, Shimomura S, et al. Flexible pneumatic twisting actuatorsand their application to tilting micromirrors[J]. Sensors and Actuators A: Physical,2014, 216: 426-431.
　　[30] Suzumori K, Endo S, Kanda T, et al. A bending pneumatic rubber actuatorrealizing soft-bodied manta swimming robot[C]//Robotics and Automation, 2007IEEE International Conference on. IEEE, 2007: 4975-4980.
　　[31] Cai Y, Bi S, Zheng L. Design and experiments of a robotic fish imitatingcow-nosed ray[J]. Journal of Bionic Engineering, 2010, 7(2): 120-126.
　　[32] Wang B, McDaid A, Biglari-Abhari M, et al. A bimorph pneumatic bendingactuator by control of fiber braiding angle[J]. Sensors and Actuators A: Physical,2017, 257: 173-184.
　　[33] Chou C P, Hannaford B. Measurement and modeling of McKibben pneumaticartificial muscles[J]. IEEE Transactions on robotics and automation, 1996, 12(1):
　　90-102.
　　[34] softroboticstoolkit.
　　[35] Galloway K C, Polygerinos P, Walsh C J, et al. Mechanically programmable bendradius for fiber-reinforced soft actuators[C]//Advanced Robotics (ICAR), 201316th International Conference on. IEEE, 2013: 1-6.
　　[36] Polygerinos P, Wang Z, Overvelde J T B, et al. Modeling of soft fiber-reinforcedbending actuators[J]. IEEE Transactions on Robotics, 2015, 31(3): 778-789.
　　[37] Ogden R W. Non-linear elastic deformations[M]. Courier Corporation, 1997.
　　[38] Polygerinos P, Lyne S, Wang Z, et al. Towards a soft pneumatic glove for handrehabilitation[C]//Intelligent Robots and Systems (IROS), 2013 IEEE/RSJInternational Conference on. IEEE, 2013: 1512-1517.
　　[39] Yeoh O H. Some forms of the strain energy function for rubber[J]. RubberChemistry and technology, 1993, 66(5): 754-771.
　　[40] Gorissen B, Reynaerts D, Konishi S, et al. Elastic Inflatable Actuators for SoftRobotic Applications[J]. Advanced Materials, 2017.
　　[41] Whitesides, George M., and Abraham D. Stroock. "Flexible methods formicrofluidics." Phys. Today 54.6 (2001): 42-48.
　　[42] Guo W, Sheng X, Liu J, et al. Towards zero training for myoelectric control basedon a wearable wireless semg armband[C]//Advanced Intelligent Mechatronics (AIM), 2015 IEEE International Conference on. IEEE, 2015: 196-201.