外文翻譯--基于黑板模式整合多種領(lǐng)域達(dá)到戰(zhàn)略上為機(jī)械手導(dǎo)航

1、 畢業(yè)設(shè)計(jì)(論文)外文資料翻譯系部： 機(jī)械工程系 專 業(yè)： 機(jī)械工程及自動(dòng)化 姓 名： 學(xué) 號(hào)： (用外文寫)外文出處： ramiro liscano, allan maw, and elizabeth r. stuck, national research council reda e. fayek, university of waterloo jean-yves tigli, universite de nice 附 件： 1.外文資料翻譯譯文；2.外文原文。 指導(dǎo)教師評(píng)語： 簽名： 2009 年 3 月 18 日注：請(qǐng)將該封面與附件裝訂成冊(cè)。附件1：外文資料翻譯譯文基于黑板模式整合多種

2、領(lǐng)域達(dá)到戰(zhàn)略上為機(jī)械手導(dǎo)航發(fā)展一個(gè)移動(dòng)的機(jī)械手，在智能系統(tǒng)設(shè)計(jì)中給予我們一個(gè)調(diào)查機(jī)會(huì)，因?yàn)闄C(jī)械手的可動(dòng)性使我們處理許多不可預(yù)知的形式環(huán)境。智能的一種字典定義是有能力處理新的嘗試情形。因此，可靠的導(dǎo)航機(jī)械手在未知的環(huán)境中提供智能表現(xiàn)，主要注意一輛自治的車輛相當(dāng)簡(jiǎn)單：給予一個(gè)操作系統(tǒng)，它一定有能力得到適當(dāng)?shù)氐沫h(huán)境和行為。這能力需要一個(gè)回給控制系統(tǒng)，5個(gè)測(cè)知和控制聯(lián)編車輛。不幸地是機(jī)械手有些特性仍然不滿足被傳統(tǒng)行為控制：遇到解決可動(dòng)裝置機(jī)械手的問題，通常需要一些方法學(xué)的整合。系統(tǒng)在一個(gè)適當(dāng)?shù)臅r(shí)期中對(duì)環(huán)境產(chǎn)生反應(yīng)。由于感應(yīng)器的限制和控制處理。大部份的知識(shí)機(jī)械手獲得的是不完全的或不確定的問題去處理。我們

3、已經(jīng)用可動(dòng)裝置機(jī)械手系統(tǒng)建筑學(xué)對(duì)同等的人物和一些即時(shí)的活動(dòng)擬一個(gè)黑板模式?；顒?dòng)領(lǐng)域是一個(gè)組織的單位或組件運(yùn)行一個(gè)特定的功能，如穿越一個(gè)走廊，下降階段，一個(gè)開著的通風(fēng)槽在底板上，或指一個(gè)陸標(biāo)，一個(gè)活動(dòng)，相似控制一個(gè)特定的操作中的行為。它被設(shè)計(jì)在一個(gè)狹窄的應(yīng)用領(lǐng)域中運(yùn)行，操作上不同于行為，然而行為通常相似一個(gè)回應(yīng)。payton 定義期間，活動(dòng)當(dāng)做活化的設(shè)定。被設(shè)定的活動(dòng)由一些行為的地方組成。payton 的活動(dòng)是敘述一個(gè)行為的結(jié)合，達(dá)成一個(gè)更復(fù)雜的動(dòng)作模型的方法。在差別中，我們沒有嘗試定義我們的活動(dòng)為一個(gè)基本行為的結(jié)合。在我們的系統(tǒng)中，一些活動(dòng)是必需的，讓機(jī)械手操作運(yùn)行簡(jiǎn)單，如此的移動(dòng)相似移動(dòng)一個(gè)

4、工廠區(qū)。這些基本的導(dǎo)航活動(dòng)正在穿越一個(gè)開著的空間。橫越底板上方偏差（纜或通風(fēng)槽），而且避免碰撞。系統(tǒng)建筑學(xué)定義一個(gè)機(jī)械裝置協(xié)調(diào)移動(dòng)機(jī)械手活動(dòng)，自從他們之后不能全部同時(shí)地駕駛機(jī)械手。大多數(shù)的可動(dòng)裝置機(jī)械手控制系統(tǒng)是結(jié)合混合系統(tǒng)的方法，從階層的，動(dòng)作的，和以黑板模式為基礎(chǔ)的系統(tǒng)。以行為為基礎(chǔ)的系統(tǒng)最近已經(jīng)變得更普遍的應(yīng)用在控制移動(dòng)機(jī)械手。階層的建筑，像是 nasrem（美國(guó)航空暨太空總署 nist 標(biāo)準(zhǔn)的功能建筑學(xué)控制系統(tǒng)建筑學(xué)）和 imas （智能的可動(dòng)裝置自治的系統(tǒng)）提議一個(gè)好的斷落使一個(gè)全球的操作進(jìn)入次操作，但是階層的結(jié)構(gòu)很快被分配測(cè)知的進(jìn)入和階層的指令結(jié)構(gòu)結(jié)合而降低，類似那在 mobot

5、3 上實(shí)現(xiàn)使用大部分系統(tǒng)，感覺想作用難以在真正的時(shí)期實(shí)現(xiàn)，當(dāng)機(jī)械手處理不同的測(cè)知狀態(tài)的時(shí)候。已經(jīng)示范的nasrem 風(fēng)格使用建筑學(xué)為即時(shí)定義，狹窄部份操作的范圍，而且不需要測(cè)知一個(gè)不同的移動(dòng)機(jī)械手。試圖把反身的能力加入一階層的系統(tǒng)， payton 連同更傳統(tǒng)的水平階層的分解，一起計(jì)劃了一個(gè)垂直的分解，造成被組成的階層結(jié)構(gòu)反身行為在最低的水平。那效果是有一個(gè)階層的能力系統(tǒng)，管理不同的測(cè)知狀態(tài)和一個(gè)比較強(qiáng)健的系統(tǒng)。 arkin也強(qiáng)調(diào)數(shù)據(jù)的非階層廣播的重要， arkin 的選擇使世界的操縱被統(tǒng)一表現(xiàn)，基于潛在的田地設(shè)計(jì)一個(gè)系統(tǒng)。我們的系統(tǒng)不使用被表現(xiàn)統(tǒng)一協(xié)調(diào)活動(dòng)，但是仰賴程序的知識(shí)，而且知覺的數(shù)據(jù)張

6、貼在黑板模式上作決定。類似怪人系統(tǒng)建筑學(xué)過去一直駕駛著林肯，我們的系統(tǒng)不以它的黑板模式作為問題，解決的機(jī)械裝置除了主要地當(dāng)做管理者和協(xié)調(diào)者和一些真正的計(jì)時(shí)活動(dòng)。這些活動(dòng)不斷地郵寄他們現(xiàn)在的州對(duì)黑板模式的環(huán)境。那知覺和測(cè)知組成物活動(dòng)被設(shè)計(jì)地協(xié)同運(yùn)行。不是活動(dòng)之間的所有組件溝通，如此可能減少帶寬溝通而且讓行為反動(dòng)。喜歡一個(gè)傳統(tǒng)的黑板模式系統(tǒng)，我們的系統(tǒng)使用一個(gè)中央的數(shù)據(jù)庫(kù)儲(chǔ)存數(shù)據(jù)，可借著一些組件，但是它的功能差異因?yàn)榻M件合作，不要費(fèi)力的去解決一般問題。在我們的系統(tǒng)建筑學(xué)中，黑色板的尺組和知識(shí)決定哪一個(gè)活動(dòng)控制移動(dòng)機(jī)械手的引動(dòng)器。策略是處理知覺的能力，儲(chǔ)存了數(shù)據(jù)。而且長(zhǎng)期的目標(biāo)以便機(jī)械手能作出全球環(huán)

7、境的決定視野，那決定哪一個(gè)有關(guān)活動(dòng)將會(huì)控制以車輛為基礎(chǔ)知覺和活動(dòng)的形成，因此是一個(gè)策略推論的形狀。可動(dòng)裝置機(jī)械手系統(tǒng)建筑適當(dāng)系統(tǒng)的疑問，給一輛自治的車輛控制是持續(xù)的研究問題，而且只有在多數(shù)的系統(tǒng)設(shè)計(jì)測(cè)試之后被解決。這研究是它幼年的劇照，因此沒有通常的協(xié)議，在哪一個(gè)系統(tǒng)建筑學(xué)是最適當(dāng)?shù)?。因?yàn)榭刂埔粋€(gè)移動(dòng)的機(jī)械手，幾乎所有現(xiàn)有的系統(tǒng)不斷地接受，作為移動(dòng)的機(jī)械手的變化，是以較多的能適合。雖然多數(shù)研究學(xué)會(huì)和一些公司是追求這研究，他們的目標(biāo)不是所有都一樣的。重要區(qū)別是在研究進(jìn)入以移動(dòng)機(jī)械手作為工具才深入的研究設(shè)計(jì)自治的可動(dòng)裝置機(jī)械手。一個(gè)系統(tǒng)想要運(yùn)行一個(gè)功能為一個(gè)工具，是發(fā)信號(hào)當(dāng)做一個(gè)傳統(tǒng)的控制系統(tǒng)，包

8、含一個(gè)計(jì)劃組件如自治的可動(dòng)裝置機(jī)械手，另一方面，應(yīng)該在不同環(huán)境采用不同的策略和機(jī)械裝置解決問題。因此，我們的行為與設(shè)計(jì)系統(tǒng)建筑學(xué)有能力整合一些可靠的功能操作。雖然研究移動(dòng)機(jī)械手工程還沒有達(dá)到成熟，但我們能列出一組可動(dòng)裝置機(jī)械手控制的系統(tǒng)屬性，我們已經(jīng)把這些屬性分為二個(gè)組：屬性描述需要移動(dòng)機(jī)械手的行為和控制一個(gè)成功設(shè)計(jì)建筑學(xué)。這些其中的一些已經(jīng)被列出由先前而且已經(jīng)被其他系統(tǒng)建筑學(xué)當(dāng)做辯護(hù)，但是沒有正式的方法已經(jīng)被采用在機(jī)械手工程社區(qū)。行為是一個(gè)可動(dòng)裝置機(jī)械手對(duì)一種刺激的回應(yīng)。我們已經(jīng)證明六種重要行為屬性對(duì)一個(gè)移動(dòng)的機(jī)械手。l 反應(yīng)度：因?yàn)檎鎸?shí)的環(huán)境是無社會(huì)組織的，移動(dòng)機(jī)械手能做關(guān)于它的少數(shù)假定動(dòng)

9、力學(xué)。它對(duì)一方面的突然改變產(chǎn)生一定的反應(yīng)指在一定的環(huán)境時(shí)間之內(nèi)。反動(dòng)行為通常明確定義作用在回應(yīng)規(guī)劃測(cè)知的領(lǐng)域，這給那機(jī)械手持續(xù)運(yùn)動(dòng)在非常有限制的范圍之內(nèi)。在我們的系統(tǒng)中所有的活動(dòng)中是反動(dòng)行為的例子。2 智能：意指那能應(yīng)付新的嘗試情形。一個(gè)機(jī)械手的能處理不同的情形，而且操縱那達(dá)成一個(gè)目標(biāo)。雖然真實(shí)的智能仍然沒被了解，所有的可動(dòng)裝置機(jī)械手計(jì)劃努力達(dá)到一些智能的行為。3 集中的全球推論：全球，高階層的決策組件是讓機(jī)械手了解它的全部情形，這是特別重要的。當(dāng)許多獨(dú)立的活動(dòng)關(guān)系，正在嘗試控制系統(tǒng)，希望與全球的視野能擔(dān)任在活動(dòng)之中的一個(gè)仲裁人4 改良系統(tǒng)的運(yùn)轉(zhuǎn)。系統(tǒng)的區(qū)別特征是全球的推論是否來自集中表現(xiàn)或合

10、作在分配之中組件。當(dāng)使用一個(gè)全球的推論，考慮一個(gè)移動(dòng)機(jī)械手穿越一個(gè)狹窄的走廊和發(fā)現(xiàn)一個(gè)物體已經(jīng)阻塞它的方法。大多數(shù)的反動(dòng)航行運(yùn)導(dǎo)航則會(huì)讓機(jī)械手停止，折回，或從邊踱步偏袒等候一次又一次的。借由全球理解情形，機(jī)械手會(huì)試著找另外的一條路徑或推物體出來。所有的替代選擇需要知道其他的來源，整合進(jìn)入之內(nèi)統(tǒng)一表現(xiàn)。5 復(fù)式目標(biāo)分解：對(duì)于移動(dòng)膚蠅的幼蟲，情形需要改變，現(xiàn)在的作用是不可避免的，系統(tǒng)應(yīng)該提供多個(gè)方法實(shí)現(xiàn)目的。一個(gè)例子，可能發(fā)生沖突的物體而避免沖突其那物體。6 ?。阂粋€(gè)系統(tǒng)的強(qiáng)健是它有處理前任事件的缺點(diǎn)能力，不確定和突然故障。這在一個(gè)真實(shí)的操作環(huán)境中是一個(gè)決定性的個(gè)性系統(tǒng)?？煽啃裕阂粋€(gè)系統(tǒng)的可靠性是

11、根據(jù)它的能力確定一定的操作或運(yùn)轉(zhuǎn)降格在特定時(shí)間內(nèi)。航行系導(dǎo)航該維持在一個(gè)持續(xù)的水平，勝任而且同樣地運(yùn)行在一些不同的環(huán)境。為例，我們會(huì)期待那碰撞避免運(yùn)算法則會(huì)操作在一些可靠地環(huán)境中和任何的運(yùn)轉(zhuǎn)降格，可能使系統(tǒng)戲劇性的改正，大部份的情形下，誤解感應(yīng)器數(shù)據(jù)引起運(yùn)轉(zhuǎn)降格，我們能改正較多的感應(yīng)器或改良那處理感應(yīng)器的運(yùn)算法則。設(shè)計(jì)規(guī)例：屬性是關(guān)于可動(dòng)裝置機(jī)械手系統(tǒng)建筑學(xué)對(duì)折的一疊紙。除了第五個(gè)屬性以外在那目錄之后，他們需要傳統(tǒng)軟件工程學(xué)，不像大多數(shù)的傳統(tǒng)軟件工程學(xué)計(jì)劃，從一開始指定需求組。我們又證明了下列的設(shè)計(jì)屬性：模組化：在對(duì)折的一疊紙大多數(shù)的復(fù)雜系統(tǒng)，那自治車輛的控制建筑學(xué)應(yīng)該被區(qū)分為較小的能實(shí)現(xiàn)的設(shè)

12、計(jì)系統(tǒng)，而且分開地去除。模組化也是逐漸增加的設(shè)計(jì)，維護(hù)、損壞發(fā)現(xiàn)和修正?？蓳闲裕簩?shí)驗(yàn)的機(jī)械手工程，在一般以感應(yīng)器為基礎(chǔ)的個(gè)別項(xiàng)目中控制，需要連續(xù)的落實(shí)在設(shè)計(jì)期間，變更逐步運(yùn)行。易曲的控制結(jié)構(gòu)是設(shè)計(jì)指導(dǎo)成功或設(shè)計(jì)機(jī)械的重要要素。模組的設(shè)計(jì)向可撓性的第一個(gè)階段是在柔軟中留心設(shè)計(jì)。伸延性：因?yàn)樗怯?jì)時(shí)設(shè)計(jì)建立，而且是測(cè)試控制系統(tǒng)的守寡組成物，可擴(kuò)張建筑學(xué)是令人想要的?？蓴U(kuò)張建筑學(xué)在多數(shù)的熟練機(jī)械手中，需要應(yīng)付不同的情形。適合性：每個(gè)活動(dòng)一個(gè)策略被控制統(tǒng)治。以后的世界是不可預(yù)知的快速發(fā)展，控制系統(tǒng)肯定在控制之間快速地適應(yīng)現(xiàn)在的情形。附件2：外文原文（復(fù)印件）using a blackboad toln

13、tegrate multiple activitiesand achieve strategicreasoning for mobile-robotnavigationd eveloping a mobile robot gives us the opportunity to investigate issues in the design of intelligent systems because the robots mobility forces us to deal with many unpredictable environmental situations. one dicti

14、onary definition of intelligence is the ability to deal with new or trying situations .thus, a mohole robot that reliably navigates in unknown environments gives the appearance of intelligent behavior. the main idea of an autonomous vehicle is quite simple: given a task to perform, it must have the

15、ability to practice the environment and act appropriately .this ability requires a feedback control system to link the vehicle.5 sensing and coition.unfortunately. autonomolis robots have characteristics not yet satisfactorily addressed by the classical control community:solving the problems encount

16、ered by the mobile robot generally requires the integration of several methodologies.the robots decision space is discrete and composed of distinct elements as opposed to continuous functions.the system must react to the environment in an appropriate time period.due to the limitations of the sensors

17、 and censor processing. most of the knowledge the robot acquires is either incomplete or uncertain.to address these problems. we have designed a mobile-robot system architecture that uses a blackboard to coordinate and integrate several real-time activities. an activityis an organizational unit. or

18、module. designed to perform a specific function, such as traversing a hallway. going down steps. crossing over an open channel on the floor, or tracking a landmark. an activity resembles a behavior in that it controls the robot to perform a specific task. it differs from a behavior in that it is des

19、igned to perform the specific task in a narrow application domain, whereas a behavior generally resembles a biological response-that is. an organisms response to a stimulus.payton defined the term activity as an instance of an activation set. where an activation set is composed of a number of behavi

20、ors. paytons activities are a way to specify a combination of behaviors to achieve a more complex behavioral pattern. in contrast, we make no attempt to define our activities as a combination of basic behaviors. in our system, several activities are necessary for the robot to perform simple tasks su

21、ch as moving around a factory bay. some of these basic navigation activities are traversing open space. crossing over floor anomalies (cables or channels), and avoiding collisions.the system architecture must define a mechanism to coordinate the mobile robots activities since they cannot all drive t

22、he robot simultaneously. most mobile-robot control systems are hybrid systems combining approaches from hierarchical. behavial,and blackboard-based systems. behavior-based systems have recently become more prevalent for controlling mobile robots.hierarchical architectures, such as nasrem (nasa/nist

23、standard functional architecture for telerobot control system architecture)j, and imas (intelligent mobile autonomous system).” offer a nice paradigm breaking down a global task into sub tasks, but the hierarchical structure quickly degrades into a hierarchical command struc combined with distribute

24、d sensing sim to that implemented on mobot iii. most these systems use sense-think-act cycles that are difficult to implement in real time when the robot must deal with diverse sensing conditions. lumia” has demonstrated the use of nasrem-stvle architecture for real-time defined, narrow scope of ope

25、ration and does not require the sensing diversity that a mobile robot needs.in an attempt to add reflexive ability to a hierarchical system, paytoni proposed a vertical decomposition along with the more classical horizontal, hierarchical decomposition, resulting in a hierarchical structure composed

26、of reflexive behaviors at the lowest level. the effect is a hierarchical system capable of managing diverse sensing conditions-andthere for a more robust system. arkin also emphasized the importance of a nonhierarchical broadcast of information. arkin chose to design a system using agents that manip

27、ulate a unified representation of the world on potential fields. our system does not use a unified representation to coordinate activities but relies on procedural knowledge and sensory data posted on the blackboard to make a decision.similar to the codgers system architecture used to drive cmus nav

28、lab,” our system does not use its blackboard as a problem solving mechanism but primarily as the supervisor and coordinator of several realtime activities. these activities continually post their current state and the current state of the environment to the blackboard. the perception and sensing com

29、ponents of an activity are designed to run concurrently.not all communications between an activitys modules go through the blackboard. thus reducing communications bandwidth and making reactive behavior possible. like a traditional blackboard system, our system uses a central database to store infor

30、mation accessible by a number of modules, but it differs functionally because the modules do not work cooperatively to solve a common problem.in our system architecture, the blackrem boards rule set and knowledge determine which activity controls the mobile robots actuators. a production system faci

31、litates experimental determination of adequate conforditions for selecting the activity controlling the vehicle, and the blackboard database serves as a repository of state data and senturesory data from the activities. strategic reailarsoning is the ability to process sensory inofputs, stored infor

32、mation. and long-termgoals so that the robot can make decisions with a global view of the environment. the decision about which activity will control the vehicle is based on sensory and state in based formation from the activities and therefore is a form of strategic reasoning.behavior attributes. a

33、 behavior is a mobile robots response to a stimulus from the environment. we have identified six behavior attributes important to a mobile robot: reactivity: because the real-world environment is unstructured, the mobile robot can make few assumptions about its dynamics.it must react to sudden chang

34、es inthe environment within a specified timeframe. reactive behavior is generally a well-defined action in response to a narrow domain of sensing, which gives the robot a very limited scope of understanding.in our system all activities are examples of reactive behavior.l inteelligence: by intelligen

35、ce we mean the ability to cope with new or trying situations a robots ability to manage diverse situations and manipulate the environment to achieve a goal gives the appearance of intelligence. although true intelligence is yet to be understood, all mobile-robot projects strive to achieve some intel

36、ligent behavior.centralized global reasoning: a global,high-level decision-making module is crucial for the robot to understand itsover all situation. this is particularly important when many independent activities, each with a narrow understanding of the situation, are trying to control the system.

37、 areasoning agent with a global view can act as an arbiter among the activitiesto improve the systems performance. a distinguishing feature of the system is whether the global reasoning comes from a centralized, uniform representation or from cooperation among distributed modules. as an example of t

38、he use of global reasoning, consider a mobile robot traversing a narrow hallway and discovering that an object has blocked its way. most reactive navigation algorithms would have the robot stop, turn back, or pace from side to side waiting for an opening. with a global understanding of the situation, the robot could try to find another path or push the object out of the way. all the