机械臂动力学外文翻译@中英文翻译@外文文献翻译_第1页
机械臂动力学外文翻译@中英文翻译@外文文献翻译_第2页
机械臂动力学外文翻译@中英文翻译@外文文献翻译_第3页
机械臂动力学外文翻译@中英文翻译@外文文献翻译_第4页
机械臂动力学外文翻译@中英文翻译@外文文献翻译_第5页
已阅读5页,还剩32页未读 继续免费阅读

下载本文档

版权说明:本文档由用户提供并上传,收益归属内容提供方,若内容存在侵权,请进行举报或认领

文档简介

毕业 论文 ( 设计 ) 外文翻译 题 目 机械臂动力学与控制的研究 系 部 名称 : 机械工程 系 专业班级: 机自 学生姓名: 学 号: 指导教师: 教师职称 : 20*年 03月 20日 1 2009 年 IEEE 国际机器人和自动化会议 神户国际会议中心 日本神户 12-17,2009 机械臂动力学与控制的研究 拉 斯彼得 Ellekilde 摘要 操作器和移动平台的组合提供了一种可用于广泛应用程序 高效灵活的操作系统 ,特别是在服务性机器人领域。 在 机械臂 众多 挑战中其 中之 一是确保机器人在潜在的动态环境中安全工作控制系统的设计。在本文中 ,我们将介绍移动机械臂用动力学系统方法被控制的使用方法。该方法是一种二级方法 , 是使用竞争 动力学对于统筹协调优化移动平台以及较低层次的融合避障和目标捕获 行为 的方法 。 I 介绍 在过去的几十年里大多数机器人的研究主要关注在移动平台或操作系统,并且在这两个领域取得了许多可喜的成绩。今天的新挑战之一是将 这两个领域组合在一起形成具有高效移动和有能力操作环境的系统。特别是服务性机器人将会在这一方面系统需求的增加。 大多数西方国家的人口统计数量显示需要照顾的老人在不断增加 ,尽管将有很少的工作实际的支持他们。这就需要增强服务业的自动化程度,因此机器人能够在室内动态环境中安全的工作是最基本的。 2 图、 1 一台由 赛格威 RMP200和轻重量型 库卡机器人 组成的平台 这项工作平台用于如图 1所示 ,是由一个 Segway与一家机器人制造商制造的 RMP200轻机器人。其有一个相对较小的轨迹和高机动性能的平台使它适应在室内环境移动 。库卡工业机器人具有较长的长臂和高有效载荷比自身的重量 ,从而使其适合移动操作。 当控制移动机械臂系统 时 ,有一个选择是是否考虑一个或两个系统的实体。在参考文献 1和 2中是根据 雅可比理论 将机械手末端和移动平台结合在一起形成一个单一的控制系统。另一方面 ,这项研究发表在 3和 4,认为它们 在设计时 是独立的实体 ,但不包括两者之间的限制条件 ,如延伸能力和稳定性。 这种控制系统的提出是基于动态系统方法 5,6。它分为两个层次, 其中 我们在较低的水平,并考虑到移动平台作为两个独立的实体,然后再以安全的方式结合 在上层操纵者 。 在本文中主要的研究目的是展现动力系统方法可以应用于移动机械臂和使用各级协调行为的控制。 本文剩下的安排如下 。第二部分介绍系统的总体结构设计 ,其次是机械手末端移动平台的控制在第三第四部分讲述。在第五部分我们 在结束本文之前将显示 一些 实验 。然而 , 首先 与 动力学系统 有关 工作总结与方法 将在 在部分 I-A提供。 3 A.相关工作 动力学系统接近 5, 6为控制机器人提供一套 动作 的 框架 ,例如障碍退避和目标捕捉。 每 个动作 通过一套一个非线性动力学系统的 attractors和 repellors来完成 。 这些通 过向量场的简单的加法被结合 在一起来完成 系统 的整体 动作 。动力系统的方法涉及到更广泛的应用势场法 7,但具有一定的优势。 这里 势场法的行为 是 由 后场梯度 形成的结果,行为变量,如航向和速度,可直接 运用 动力系统 控制 的方法。 成本相对较低的计算与方法有关,使得它在动态环境中在线控制适宜,允许它即使在相当低的水平有限的计算能力平台 8实施。传感器的鲁棒性 在 人声嘈杂中显示 9和10其中一个 是由 红外传感器和麦克风的结合, 当 避障和目标获取 时 使用。尽管能解决各种各样的任务 ,但 它 仅是 一个 局部的 方法,为了 其他的任务 和使命级 计 划 (即参见 11)其他的方法 应该 被采用 。 当多行为被结合时 ,在 5和 6的缺点是由潜在的假的因子引起的。 为了克服这个问题 12介绍了一种基于 竞争动态的行为比重。每个行为的影响是控制使用一个 相关的竞争优势,再加上定义的行为之间有竞争力的相互作用,控制 重物 。 如果所有的行为之间的竞争性相互作用是必 需 的 , 这种方法可以推广到任意数 n,行为, 除了这样一个 最坏情况 的 复杂 度 2n 。 在 现实世界中使用这种方法的竞争态势室内实验中可以找到 13, 14。 13是 只在 有标题方向的车辆 上 使用,而在 14中 航向和速度 均 得到控制。 15提供了一个 为 速度 性能 简短的 策略 讨论 。 在 16中提到 动力系统的方法不仅被用于平面移动机器人,同时也 可以作 为控制机械手工具 。 另外运用 产生极限环 Hopf振荡器动力系统 的 更复杂的 动力系统 也可 被 使用。 17展现出 不同形状的极限环 是如何产生的 , 其可运用于避障轨迹的生成 。 18中介绍到 使用 Hopf振荡器产生一个定时的轨迹,实现了机械手 可以 接 住从桌子上面滚下来的球 。动力系统的方法不仅可以用于控制的工具,也 可以 控制 7 自由度 机械手 多余的动作这 一点在 19中得到论证 。 II.总体结构 我们 整个 系统的整体架构如图 2 所示。 在 赛格威平台 中为了 控制移动 平台 ,两个低级别的 性能被 使用:一个用于目标捕获和 另一个是 避障。 运用竞争动态的动作被混合在 4 一起是为了做出移动平台希望得到的指定的移动动作 。同样, 在 竞争态势的基础上 目标捕获和机械手避障行为 的 融合给机器人 收缩下达指令 。当目标不在范围 内 ,应收回机械手到一个安全的 位置 ,这是机械手缩回行为的目的。最后融合 是以一 个安全的 方式把所有的控制结合在一起 ,这样 一来 目标捕获和收回行为不互相干扰 ,另外 移动平台 在 不开始朝着新的目标之 前,移动机械手已被收回。 图 .2. 控制系统的 体系结构 用 wmobile 、 manipacquisitionw和 manipretractw分别 代表 机械手 移动 、 机械手 捕获 和机械手 收缩 行为的影响,控制信号 mobileu 和 manipq 通过( 1)( 2) 移动平台和机械手 。 le ftr ig h tum o b ile m o b ile uuw ( 1) m a n i p m a n i pm a n i p m a n i pm a n i pa c q u i s i t i o na c q u i s i t i o n r e t r a c t r e t r a c tqq qww (2) 其中(leftu rightu) 是指 控制输入 信号以控制 在第三节中描述的平台的左,右侧车轮 ;manipacquisitionq 和 manipretractq 是 在第四节描述的机械手关节速度。 A.竞争动态 5 这种 竞争态 势采用的方法是 以 12为基础 的 , 除了附加 参数bT用于控制在 14中 的转换率。动力系统采用 ( 3) 因此给予 : 3 2( ) ,b b b b b bbbbT w a w w r b b w w n o i s e ( 3) 其中ba是 b和 rb 竞争优势 产生的参数 , b是 b 和 b相互 竞争 作用的参数 。 1)移动: 在移动平台 远离目标时它的竞争优势应该被加强 ;当目标被捕获时移动平台的竞争优势应该被降低。 这是通过 ( 4)实现的。 t a n h ( ( ) )m o b i l e m o b i l e m o b i l ea t a r t h r e s h o l da k d d ( 4) 其中, mobileak决定如何迅速的 改变 这种优势 ,tard是 指 到目标的距离和 mobilethresholdd是 指 移动平台移动目标 所需的最小距离 。 移动的行为,没有能力进行互动,并抑制其他行为,因此它的竞争性相互作用被设置为 0。 2) 机械手 捕获目 标 : 当移动平台接近他的目标时,机械手捕获目标的动作应该别加强 。这样的竞争优势将被定义为 : t a n h ( ( ) )m a n i p m a n i p m a n i pa c u i s i t i o n a t a r t h r e s h o l da k d d ( 5) 激活距离 manipthresholdd必须大于 mobilethresholdd来确保其行为被激活。此 动作没有和其他的动作有直接 联系 ,因此它的 相互作用参数 设置为 0。 3)机械手收缩: 收回 动作 应该被激活 当对面目标被捕获之后 ,因此 m anip m anipretract acqisitionaa t a n h ( ( ) )m a n i p m o b i l ea t a r t h r e s h o l dk d d ( 6) 要有 一个非常小的过渡时间 , 这可以防止在同一时间活动的机械臂 捕获 和 收缩动作 ,因此,我们可以设 置, 0retract acquisitionr 。由于机械手收缩和移动动作的联系,当机械手原理自动巡航装置时我们希望能够取消停止移动 。 因此 这种相互作用 定义为 : hom, 1 ( 1 t a n h ( ( ) ) )2 r e t r a c tr c u r r e n t e qr e t r a c t a c q u i s i t i o n k q qr ( 7) 6 其中curq和homeq,是 机械手 当前和原始配置参数 ,q是 指 目标homeq最近 的距离和 retractrk指定如何 使 相互作用 迅速变化 的参数 。 III. 移动平台的控制 该 移动平台的控制, 结构 与参考文献 14中表述的 非常相似 ,但 也 有一些 不同 。 刚开始时 目标捕获和避障 指令被使用 。 紧接着除 走廊和墙壁 避障 不包括在内,但将 沿直线扩展。第二个领域,不同的是这项工作的障碍 是 如何 找出障碍 密度的计算方法。 具体的论述 在 III-D部分 。 为了使控制系统能够根据具体的环境进行导航 。我们所使用的方法是基于 参考文献20中论述 的方法,它 运用 里程计和激光测距 相结合 对 所 在环境 中 地图 的主导线 匹配测量。 该平台 控制编码的使用方向 : ; 速度 : V,它在一个控制输入系统的结果 数 ,m obilef 的值是由两部分组成, mobiletarf 和 mobileobsf ,这 里 合并为 m o b i l e m o b i l e m o b i l em o b i l e m o b i l et a r o b st a r o b sf f fww ( 8) 其中 mobiletarw和 mobileobsw是被 Eq限制 的。 (3)中的 竞争优势 和相互作用在 III-C中有详细的描述 。 作为控制输入我们需要一个表达式对移动平台的左右轮进行控制 , 这里用leftu和,rightu分别 作为 左,右侧车轮的表达 参数 。要 使 获得这些 数据 v 集成得到 v,连同所需的旋转速度 时,车轮直径wheeld和车轮之间的距离wheelbased可以用数据 库来计算控制输入 : ( , ) 2l e f tw h e e lvv du ( 9) ( , ) 2r i g h t r i g h tw h e e lvv duu ( 10) 这里 车 轮 需要 的速度差 被定义为: w heelbasew heeldd ( 12) 7 A.动态目标 : 捕获目标动作 的基本动力是 : , ,( ) s i n ( )m o b i l e m o b i l et a r t a rt a rf (13) , , m a x( ) ( m i n ( , ) )m o b i l e v m o b i l e v m o b i l et a r t a r t a rt a r v k d v vf (14) 其中 ,mobiletar 和 ,mobilevtar是吸引子的优势 参数 和tar表示运动到 目标 的 方向。 常数 mobiletark表达出机械手 到目标之间的距离和所需的速度关系。最后 最大速度maxv是 指移动平台所允许的最大速度 。 B.障碍动态 假 定一个距离,obsid, 方向参数 i 表示 机械手到第 i个障碍的 方向 ,在避障的动力学中用 公式( 15)( 16)表示如下: 22,(),2, ()im o b i l eo b s o b s i im o b i l e cdm o b i l eo b s io b s i eef (15) ,m i n m i n,m i n m a x ,m a x , m a x ,()0()m o b i l e vobsm o b i l e vio b s im o b i l e vo b s i iv v f o r v vf o r v v vv v f o r v vf (16) 其中m a x , , m i nm a x ( , )i o b s o b s iv k d v 动态 参数 包括三个要素:(一)障碍物 ()i的相对方向 ,(二) 比例系数 ,mobileobs obs icde ,其 中 mobileobsc根据距离,obsid决定 衰减的 程度。 (三) 另一个比例系数 22()2 iie 根据到 障碍 的方向而定的 ,并 运用,1a r c s i n ( )1 si o b s iDd确保两 障碍 间的 attractor 产生 ,如果机器人可以在确保安全距离 DS下 通过 。 我们可以在参考文献 14中看到具体的描述 。 对于 ,mobile vobs if是表示 调整速度 转向,obs obs ikd, 但确保 minv 最小速度 是 被保留 的 。 运用公式( 17) 获取 我们总结 所有障碍 mobileobsf的 值 : 8 ,( ) ( )m o b i l em o b i l em o b i l e o b s iobsm o b i l e v m o b i l e vobs io b s o b s ifffff ( 17) C.竞争动态 在竞争态势的 运算 如上面所述 公式 ( 3) 控制的 。下面是最大的竞争优势和两种 动作的相互作用。 1) 目标 : 每当一个目标是存在的 , 竞争优势 的参数就被 设置为tar 0.5mobile , 否则 设置为tar 0.5mobile 。 目标 动作有能力 能力 影响 和抑制避障 动作 ,目标之间的距离和最 近 的 目标之间的 比例足以确保 向目标移动的动作 是无碰撞运动。这 时 建模为 : , m i n, l i m1 ( 1 t a n h ( ( ) ) )2 obsm o b i l e m o b i l e m o b i l et a r o b s g a i n i tt a rrrdd ( 18) 其中,minobsd到最近障碍物的距离, mobilegainr是一 个如何快速是动作相互影响的 增益常数 , 我们将开始抑制避障 时lim 1mobileitr 表示 障碍和目标 之间 的距离比。 2)障碍: 该障碍 动作的 竞争优势 有公式( 19)控制: 00t a n h ( )m o b i l e m o b i l em o b i l eobs m o b i l e ( 19) 其中 mobile 是 障碍 密度 在第三节 - D被 定义。 这种相互作用被定义为 0,1 (1 t a n h ( ) ) (1 ) ) )2m o b i l e m o b i l eo b s t a r t a r o b s ( 20) 第一部分01 ( 1 t a n h ( ) )2 m o b i l e m o b i l e 抑制 目标动作当 障碍浓度超过临界值 0mobile 时 ,最后一部分,1 mobiletar obs可以确保这只是发生 在由于,mobiletarobs的原因 避障没有被抑制 。 D.障碍密度的计算 假设一系列的距离,,obsid, 移动平台和障碍的密度 ,计算 公式 为 9 ,1m a xi o b s id ( 21) 此处的定义不同于 14中的,obs idi e 。 公式化 的主要问题是,我们不能区分物体的相对多远 和 一个对象 相对多近 。例如 2米外 有 5个对象 的密度定义成 相同的密度 与 40厘米的距离 之外的一个对象 。 根据 指数函数 的性质 在场景中的单个对象永远不能导致 超 1。用于切换到避障 动作 的 临界值 将因此必须小于 1,但 一个 场景中 有多样 的障碍往往临界值设置的更低 。 此外,发现 用 ,1obsid 代替 ,obsie 参数 调整 更 容易,因为我们 可以考虑其 作为距离 的 反比密度。 这也造成了当越来越接近一个障碍 时 密度增长非常迅速,从而 可以 迅速迫使 动作改变。 IV.机械手的控制 我们将 这个问题 分成两部分 : 1) 确定 机械手的运动 ,从当前位置到目标,同时避免障碍。 2) 计算所需刀具 的 逆 运动 的 速度 。 第二部分是一个很好的理解问题,这项工作 可以运用在参考文献 23中描述的逆运动学方法 解决。这种方法包括机器人运动学和动力学的 局限性 ,如关节的位置,速度和加速度的 限制。此外, 在此 方法的基础上, 进行 二次优 化 获得方法 已 被 证明 表现很突出 。 该 机械手的运动受 机器人控制 的 目标和障碍 动作限制 , 为 此 maniptar和 manipobs是相关的 。由于 逆运动学 的输入 需要一个六维 旋转速度 , , , , ,x y zx y z ,因此这些动作必须设置一个变数 , , , , ,m a n i p x y zf x y z , 它可以集成所需的 速度 des .,( ( ) )m a n i p m a n i p m a n i p m a n i p m a n i pd e s c u r t a r t a r o b s o b s d i r o b s d i s td t f f f ( 22) 其中 maniptarf,.manipobsdirf和,manipobsdistf是从目标和 避障中 得到的 。 A.目标动作 到目标行为的输入是当前和所需的工具转换cur和des。 从这些我们可以计算出所需 10 的六维速度螺杆tar。 为避免要求不切实际的快速运动 它的范围是 m a x,ta rx y z v和 m a x,x y z ta rw w w , maxv 和 max 代表 最大允许的机床直线和旋转速度。 计算 ()m a n i p m a n i pt a r t a r c u r t a rf ( 23) 我们得到了当前 速度预期的变化。 B.障碍动作 作为输入避障 动作的参数 ,采用当前笛卡尔速度 ,v x y z , 采用 最近的障碍为 轨道 , 3inR给出机械手 和障碍物 之间方向和距离 。我们现在要 根据到障碍物的方向和距离计算笛卡尔速度的变化 ,并分别 用.manipobsdirf和,manipobsdistf表示 。 1)施力方向: 根据当前机械手的速度 V, 我们计算 向量 in 相互两者之间的 角 度 i 为 a r c s i n iiivnvn ( 24) 在 机械手尺寸 方向变化的大小, 用( 25) 计算 22, 22m a n i p io b s im a n i po b s i cnee ( 25) 其中 ,obsmanip是 repellor 的数值 , manipobsc根据距离控 制衰减 ,控制 相 对障碍 之间的 角度。 被用于计算预期的机械手方向的改变: iiivnvn ( 26) 根据所有障碍物的作用 ,我们可以 根据 障碍物的方向 计算 机械手运动的改变 : , ()m a n ipio b s d ir i vf ( 27) 2) 动力学速度 : 对速度的动态控制相似于 Eq。 障碍 i的作用是 : 11 ,m i n m i nm i n m a x ,m a x , m a x ,()0()m a n i p v m a n i p m a n i pobsm a n i pm a n i pio b s v e l im a n i p viiobsv f o rf o rv f o rf ( 28) 其中m a x ,m a x , m a x ( , )m a n i po b s i ii kn 。集合 所有障碍的 作用 变成 : ,m a n ip m a n ipo b s o b s v e l iiff ( 29) C.竞争动态 1)目标动作: 对于移动平台 当目标存在目标动作的竞争优势值设置为 0.5, 否则设置为 0.5。 当到目标的距离和最近障碍物的距离之间的比例系数超过limmanipitr, 目标与障碍物之间的相互作用需要被重新设置,避障作用受到限制, 这是 有公式( 30)实现: l i m,m i n ( )1 ( 1 t a n h ( ( ) ) )2m a n i p iim a n i p m a n i pg a i n i tt o o lt a r o b st a rnrrd ( 30) 其中 tooltard是机床和目标的距离 ; mobilegainr是一个 如何迅速改变,maniptarobs值 的 增益系数 。 2)障碍: 该 障碍动作的竞争优势和 在第三节 - C表述的 相同 : 00t a n h ( )m a n i p m a n i pm a n i p m a n i pobs ( 31) 用 Eq( 21) 进行密度计算, 但 用 障碍和 机械手 之间的距离 代替 障碍和 移动 平台 的距离 。 这种相互之间的作用用公式确定: 0o b s , ,1 ( 1 t a n h ( ) ) ( 1 )2 m a n i p m a n i pm a n i p m a n i pt a r t a r o b s ( 32) 其中当机械手最接近目标时,,(1 )maniptar obs有助于撤销臂章动作 。 D 收缩 收缩动作是在关节处直接运作的 。 通过定义,hom e c u rq qq , 其中homeq是 指机械手原始的收缩数据配置 ,我们可能计算 关节速度为 : 12 m a xm i n ,m a n i pr e t r a c t r e t r a c t m a n i pr e t r a c tqqqqq ( 33) 其中maxq是 关节 最大的速度 , manipretract为 attractor 的作用参数 。 V.实验 本实验的目的主要是展示了移动平台和机械手的协调。以前的工作已经展示了动力系统方面的方针与导航的能力通过一个环境中移动机器人 13 14和指导一个机器人绕过障碍 16。 (a)移向目标( t=0s) (b)图像伺服 (t= 28s) (c) 移动到目标 位置 (t = 40s) (d) 完成动作 (t = 72s) 图 .3移动机器人实验。 假定环境和目标重物的角度是不变的。 在实验中使用的平台 如图 1所示,是由 一个赛格威 RMP200和 轻重量型 库卡机器人 与崇德 PG70平行爪装备组成。该平台具有一个 SICK LMS291定位 和避障 装有 Unibrain 13 Fire-iFireWire摄像头 的 激光扫描仪, 用于机械手瞄准并抓起目标 。不幸的是我们没有足够的时间来连接夹持器 和控制目标 。因此, 它仅仅是定位和准备抓。 但实际上从未关闭的抓手。由于控制 框架我们 使用了 Microsoft Robotics Studio1.5,这提供了一个 从 传感器的各种输入, 到 驱动器输出,并确保不同的 控制算法同时运作的方法 。 该赛格威 运动 和 大多数 机械手运动是基于特定 的笛卡尔坐标定位目标的 。但是,一旦目标 在 toolmounted 相机 视线范围 内,机械手 依靠视觉输入 指导切换。 第五部分 A将会详细阐述视觉伺服系统方法,紧接着在第五部分 B中会提供测试结果。 图 .4.检测使用微软机器人 SimpleVision方面的服务 特征 . 黑白边边框表示 特征识别 。 A.伺服系统 对于 最终机械手的定位 是使用视觉伺服系统方法获得标准图像进行定位的 。特征 检测 是 根据 Microsoft Robotics Studio 的 SimpleVision 服务 而测定的,获得 能够识别颜色的斑点。在这些 试验中获得结果我们用 绿色标记 标出 ,如图 4所示。我们希望该 机械手 的方向是固定的,因此 仅仅需要 3个自由度(自由度)的 位置 应该 被 相关的视觉输入的影响。这些自由度两个是 由 BLOB的 定位控制 , 其中一个 应在图像中心位置。最后的自由度是由 BLOB的大小 决定的 。 B.测试结果 如 图 3 所示 , 移动机械手的任务是 移动一个瓶子从图像的桌子上移动到右边相对 的较远的箱子里。 机器人 移动、 机械手 收缩和 目标行为有关的 数据关系 可以在图 5中看到 。 14 图 .5 机械手运行时各项的 比例系数表 首先 移动机械手 收缩 和 移动指令被激活引起移动 平台 移 向目标,同时 手臂 保持原始的配 置 装态 。经过约 7秒之内达到目标 并 获得 目标信号 ,因此 机械手收缩动作被取消,机械手捕获动作 被激活。不久后, Segway动作 也 被取消 ,让机械手拿起无干扰的 目标 。然而 机械手运动 会 导致赛格威漂移, 因此要过一会知道经过 20s之后移动平台重新被激活 ,在 这 里 移动平台 又达到了预期目标的相对位 置 。视觉伺服 指挥机械手到 如图 3( b)所 示的状态 。经过约 30秒钟,瓶子应该被 抓手 拾起的 和新的目标是给予,造成机械手 收缩动作被重新激活而机械手捕获动作被取消 。 同时 移动平台 移动动作 也被激活,但 当机械臂被收回时移动平台的移动动作会迅速被取消 。 完成之后 控制 移动 平台移动到所需位置放置 ,进而 机械手被激活 把目标放到箱子里 。 VI.结论 本文已经介绍了如何 使 动态系统的方法应用于移动操作。 此文的主要结论 包括两个层次,其中 竞争态势是用于移动平台的整体协调和机械手 运动 以及避障和目标获取 等动作 。该方法 首先 已被证实在模拟环境中, 其次也 通过实际工作的 验证 。 15 实验用的系统是 Microsoft Robotics Studio1.5( MSRS)。该系统最初是模拟和参数的调整,采用模拟器进行。基于模拟器的物理参数理想的转向。 整个 MSRS是一个 执行工作 有益环境 的平台 。 虽然 控制是以 20Hz 被执行的 ,但由于 Windows XP 的 非实 性 ,动作 间 会 有异常值 出现 。 本文出自 2009年 IEEE国际机器人和自动化会议 论文集 参考文献 1 H. Seraji, A Unified Approach to Motion Control of Mobile Manipulators, The International Journal of Robotics Research, Vol. 17, No. 2, 1998, pp. 107-118. 2 E. Papadopoulos, J. Poulakakis, Planning and Model-Based Control for Mobile Manipulators, Proceedings of the IROS00 , 2000, pp. 1810-1815. 3 Q. Huang, K. Tanie, S. Sugano, Coordinated Motion Planning for a Mobile Manipulator Considering Stability and Manipulation, Thee International Journal of Robotics Research, Vol. 19, No. 8, 2000, pp. 732-742. 4 D.H. Shin, B.S. Hamenr, S. Singh, M. Hwangbo, Motion Planning for a Mobile Manipulator with Imprecise Locomotion, Proceddings of the IROS03, 2003, 847-853. 5 G. Schoner, M. Dose, A dynamical systems approach to task-level system integration used to plan and control autonomous vehicle motion, Robotics and Autonomous Systems, Vol. 10, 1992, pp. 253-267. 6 G. Schoner, M. Dose, C. Engels, Dynamics of behavior: theory and applications for autonomous robot architecture. Robotics and Autonomous Systems, Vol. 16, 1995, pp. 213-245. 7 O. Khatib, Real-Time Obstacle Avoidance for Manipulators and Mobile Robots. The International Journal of Robotics Research, Vol. 5, No. 1, 1986, pp. 90-98. 8 E. Bicho, G. Schoner, The dynamic approach to autonomous robotics demonstrated on a low-level vehicle platform. Robotics and Autonomous Systems, Vol. 21, 1997, pp. 23-35. 9 E. Bicho, P. Mallet, G. Schoner, Using Attractor Dynamics to Control Autonomous Vehicle Motion. Proceedings of the IECON98, Vol. 2, 1998, pp. 1176-1181. 10 E. Bicho, P. Mallet, G. Schoner, Target Representation on an Autonomous Vehicle with Low-Level Sensors The International Journal of Robotics Research, Vol. 19, No. 5, 2000, pp. 424-447. 11 H. Choset, K.M. Lynch, S. Hutchinson, G. Kantor, W. Burgard, L.E. Kavraki, S. Thrun, Principles of Robot Motion . The MIT Press, 2005. 16 12 E.W. Large, H.I. Christensen, R. Bajcsy, Scaling the Dynamic Approach to Path Planning and Control: Competition among Behavioral Constraints. The International Journal of Robotics Research, Vol. 18, No. 1, pp. 37-58. 13 P. Althaus, H.I. Christensen, F. Hoffmann, Using the Dynamical System Approach to Navigate in Realistic Real-World Environments. Proceedings of IROS01 , Vol. 2, 2001, pp. 1023-1029. 14 P. Althaus, Indoor Navigation for Mobile Robots: Control and Representations, Ph.d. Dissertation, Royal Institute of Technology (KTH), Stockholm, Sweden, 2003. 15 S. Goldenstein, E. Large, D. Metaxas, Non-linear dynamical system approach to behavior modeling, The Visual Computer, Vol. 15, 1999, pp. 349-364. 16 I. Iossifidic, G. Schoner, Autonomous reaching and obstacle avoidance with the anthropomorphic arm of a robotics assistant using the attractor dynamics approach, Proceedings of ICRA04, 2004, pp. 4295-4300. 17 L.-P. Ellekilde, J.W. Perram, Tool Center Trajectory Planning for Industrial Robot Manipulators Using Dynamical Systems, The International Journal of Robotics Research, Vol. 24, No. 5, 2005, pp. 385-396. 18 C. Santos, M. Ferreira, Ball Catching by a Puma Arm: a Nonlinear Dynamical Systems Approach, Proceedings of IROS06, 2006, pp.916-921 19 I. Iossifidic, G. Schoner, Dynamical Systems Approach for the Autonomous Avoidance of Obstacles and Joint-limits for an Redundant Robot Arm. Proceedings of the IROS06, 2006, pp. 580-585. 20 P. Jensfelt, H.I. Christensen, Pose tracking using laser scanning and minimalistic environment models, IEEE Transactions on Robotisc and Automation, Vol. 17, No. 2, 2001, pp. 138-147. 21 J. Forsberg, P. A hman, . Wernersson, The Hough transform inside the feedback loop of a mobile robot, Proceedings of ICRA, Vol 1, 1993, pp. 791-798. 22 K.O. Arras, R.Y. Siegwart, Feature Extraction and scene interpredation for map-based nagivation and map building, Proceedings of SPIE, Mobile Robotics XII, Vol. 3210, 1997, pp. 42-53. 23 L.-P. Ellekilde, P. Favrholt, M. Paulin, H.G. Petersen, Robust control for high-speed visual servoing applications, International Journal of Advanced Robotic Systems , Vol. 4, No. 3, 2007, pp. 272-292. 17 2009 IEEE International Conference on Robotics and Automation Kobe International Conference Center Kobe, Japan, May 12-17, 2009 Control of Mobile Manipulator using the Dynamical Systems Approach Lars-Peter Ellekilde Abstract The combination of a mobile platform and a manipulator, known as a mobile manipulator, provides a highly flexible system, which can be used in a wide range of applications, especially within the field of service robotics. One of the challenges with mobile manipulators is the construction of control systems, enabling the robot to operate safely in potentially dynamic environments. In this paper we will present work in which a mobile manipulator is controlled using the dynamical systems approach. The method presented is a two level approach in which competitive dynamics are used both for the overall coordination of the mobile platform and the manipulator as well as the lower level fusion of obstacle avoidance and target acquisition behaviors. I. INTRODUCTION The majority of robotic research has in the last decades focused on either mobile platforms or manipulators, and there have been many impressive results within both areas. Today one of the new challenges is to combine the two areas, into systems, which are both highly mobile and have the ability to manipulate the environment. Especially within service robotics there will be an increased need for such systems. The demography of most western countries causes the number of old people in need of care to increase, while there will be less working to actually support them. This requires an increased automation of the service sector, for which robots able to operate safely in indoor and dynamic environments are essential. 18 Fig. 1. Platform consisting of a Segway RMP200 and a Kuka Light Weight Robot. The platform used in this work is shown in Figure 1, and consist of a Segway RMP200 with a Kuka Light Weight Robot. The result is a platform that has a relative small footprint and is highly maneuverable, making it well suited for moving around in an indoor environment. The Kuka Light Weight Robot has a fairly long reach and high payload compared to its own weight, making it ideal for mobile manipulation. When controlling a mobile manipulator, there is a choice of whether to consider the system as one or two entities. In 1 and 2 they derive Jacobians for both the mobile platform and the manipulator and combine them into a single control system. The research reported in 3 and 4, on the other hand, considers them as separate entities when planning, but do include constraints, such as reachability and stability, between the two. The control system we propose is based on the dynamical systems approach 5, 6. It is divided into two levels, where we at the lower level consider the mobile platform and the manipulator as two separate entities, which are then combined in a safe manner at the upper level. The main reesarch objective in this paper is to demonstrate how the dynamical systems approach can be applied to a mobile manipulator and used to coordinate behaviours at various levels of control. The remaining of this paper is organized as follows. The overall architecture is described in Section II, followed by the control of the mobile platform and the manipulator in Sections III and IV. In Section V we will show some experiments before concluding the paper in Section VI. However, first a summary of work related to the dynamical systems approach will be provided in Section I-A. 19 A. Related Work The dynamical systems approach 5, 6 provides a framework for controlling a robot through a set of behaviors, such as obstacle avoidance and target acquisition. Each behavioris generated through a set of attractors and repellors of a nonlinear dynamical system. These are combined through simple addition of the vector fields to provide the overall behavior of the system. The dynamical systems approach relates to the more widely used potential field method 7, but has certain advantages. Where the behavior in the potential field method is the result of following gradients of the field, the behavior variables, such as heading direction and velocity, can be controlled directly using the dynamical systems approach. The relative low computational cost associated with the approach, makes it suitable for online control in dynamic environments, and allows it to be implemented even on fairly low-level platforms with limited computational capabilities 8. The robustness to noisy sensors is shown in 9 and 10 where a combination of infrared sensors and microphones is used for obstacle avoidance and target acquisition. Despite being able to solve various tasks it is only a local method, for task and mission-level planning other methods (see e.g. 11) should be applied. A drawback of the approach in 5, 6 is the potential creation of spurious attractors when multiple behaviors are combined. To overcome this problem 12 introduces a weighting of the behaviors based on competitive dynamics. The influence of each behavior is controlled using an associated competitive advantage, which together with competitive interactions defined between the behaviors, controls the weights. This approach generalizes to an arbitrary number, n, behaviors, but with a O(n2) worst-case complexity, if competitive interactions between all behaviors are needed. Real-world indoor experiments using this competitive dynamics approach can be found in 13, 14. In 13 only the heading direction of the vehicle is used, whereas in 14 both heading direction and velocity are controlled. 15 provides a brief discussion of strategies for the velocity behavior. The dynamical systems approach has not only been used for planar mobile robots, but also for controlling the tool motion of a manipulator 16. More complex dynamical systems using the Hopf Oscillator for generating limit cycles can also be used. 17 shows how limit cycles with different shapes can be constructed and used for both obstacle avoidance and trajectory generation. 18 uses the Hopf Oscillator to generate a timed trajectory, enabling a manipulator 20 to catch a ball rolling down a table. The dynamical systems approach can not only be used for controlling the tool, but also to control the redundancy of a 7 degrees of freedom manipulator as demonstrated in 19. II. OVERALL ARCHITECTURE The overall architecture of our system is illustrated in Figure 2. To control the mobile platform, in this case a Segway, two low level behaviors are use: One for target acquisition and one for obstacle avoidance. Using competitive dynamics these are fused together to provide the Mobile behavior, which specifies the desired motion of the mobile platform. Similarly we have target acquisition and obstacle avoidance behaviors for the manipulator fused together based on competitive dynamics, to give the Manipulator Acquisition behavior. When the target is not within reach, the manipulator should retract to a safe configuration, which is the purpose of the Manipulator Retract behavior. The last fusion combines the controls in a safe manner, such that the target acquisition and retract behaviors do not disturb one another and the mobile platform does not start moving towards a new target before the manipulator has been retracted. Fig. 2. Overall architecture of the control system Using weights wmobile , manipacquisitionw and manipretractw to represent the influence of the Mobile, Manipulator Acquisition and Manipulator Retract behaviors, the control signals mobileu and manipq for the mobile platform and the manipulator are given by le ftr ig h tum o b ile m o b ile uuw ( 1) 21 m a n i pm a n i p m a n i pm a n i p m a n i pa c q u i s i t i o na c q u i s i t i o n r e t r a c t r e t r a c tqq qww (2) Where (leftu rightu) are control inputs to the left and right wheels of the platform as described in Section III, manipacquisitionqand manipretractqare the manipulator joint velocities as described in Section IV. A. Competitive Dynamics The competitive dynamics approach used is based on 12, but with the additional parameter bT used to control the transition rate as in 14. The dynamical system used is thus given by 3 2( ) ,b b b b b bbbbT w a w w r b b w w n o i s e ( 3) In which ba is the competitive advantage of behavior b and r b ,b is the competitive interaction of behavior b upon b. 1) Mobile: The competitive advantages of the mobile platform should strengthen the behavior when far away from the target and reduce it when the target is reached. This is achieved through t a n h ( ( ) )m o b i l e m o b i l e m o b i l ea t a r t h r e s h o l da k d d ( 4) In which mobileak determines how rapidly the advantage should change, tard is the distance to the target and mobilethresholdd specifies a minimum distance to the target required before the mobile platform should move. The mobile behavior has no ability to interact and suppress other behaviors, thus its competitive interactions are set to 0. 2) Manipulator Acquisition: This behavior should be strengthened when the mobile platform gets close to its target. The competitive advantage will thus be defined as t a n h ( ( ) )m a n i p m a n i p m a n i pa c u i s i t i o n a t a r t h r e s h o l da k d d ( 5) The activation distance manipthresholdd must be greater than mobilethresholdd to make sure the behavior is activated. This behavior has no direct interaction with the others, thus its interactions are set to 0. 22 3) Manipulator Retract: The retract behavior should be activated opposite the goal behavior, hence m anip m anipretract acqisitionaa t a n h ( ( ) )m a n i p m o b i l ea t a r t h r e s h o l dk d d ( 6) Except for a very small transition time this prevents the manipulators acquisition and retract behaviors from being active at the same time, thus we can set , 0retract acquisitionr . For the interaction between the retract and the mobile behaviors we wish retract to deactivate mobile when the manipulator is far away from its home configuration. The interaction is therefore defined as hom, 1 ( 1 t a n h ( ( ) ) )2 r e t r a c tr c u r r e n t e qr e t r a c t a c q u i s i t i o n k q qr ( 7) In which curq and homeq are the manipulators current and home configurations, q specifies a proximity distance around homeq and retractrk specifies how quickly the interaction changes. III. CONTROL OF THE MOBILE PLATFORM The control of the mobile platform is constructed very similar to what is presented in 14, but with a few differences. First of all only the target acquisition and obstacle avoidance behaviors are used. The corridor following and wall avoidance are not included, but would be straight forward extensions. The second area in which this work differs is in how the density of obstacles is calculated. Details of this will be explained in section III-D. For the control to actually be able to navigate through the environment, it is necessary with a method for localization. The approach we have used is based on the method described in 20, which combines odometry and laser range measurements matched against a map of dominating lines in the environment. The control of the platform is encoded using the orientation, and the velocity, , which results in a system with control inputs ,mobilefv ; The values of mobilef are made up of two parts, mobiletarf and mobileobsf , which are combined as m o b i l e m o b i l e m o b i l em o b i l e m o b i l et a r o b st a r o b sf f fww ( 8) 23 Where the weights mobiletarw and mobileobsw are controlled using Eq. (3) with the competitive advantage and interactions described in section III-C. As control input we need expressions for the left and right wheels of the mobile platform, denoted leftu and rightu , respectively. To obtain these v is integrated to get v, which together with the desired rotational velocity , the wheel diameter wheeld and the distance between the wheels wheelbased can be used to calculate the control inputs as ( , ) 2l e f tw h e e lvv du ( 9) ( , ) 2r i g h t r i g h tw h e e lvv duu ( 10) Where is the needed difference in wheel speed given by w heelbasew heeldd ( 12) A. Target Dynamics The basic dynamics of this target behavior is , ,( ) s i n ( )m o b i l e m o b i l et a r t a rt a rf (13) , , m a x( ) ( m i n ( , ) )m o b i l e v m o b i l e v m o b i l et a r t a r t a rt a r v k d v vf (14) In which ,mobiletar and ,mobilevtar are the strengths of the attractors and tar is the direction to the target. The constant mobiletark gives the relation between the distance to the target and the desired velocity. Finally maxv is the maximal velocity allowed for the mobile platform。 B. Obstacle Dynamics Given a distance ,obsid and a direction i to the ith obstacle, the dynamics of the obstacle avoidance are 22,(),2, ()im o b i l eo b s o b s i im o b i l e cdm o b i l eo b s io b s i eef (15) 24 ,m i n m i n,m i n m a x ,m a x , m a x ,()0()m o b i l e vobsm o b i l e vio b s im o b i l e vo b s i iv v f o r v vf o r v v vv v f o r v vf (16) Where m a x , , m i nm a x ( , )i o b s o b s iv k d v The dynamics of consists of 3 elements: (i) The relative direction to the obstacle ()i , (ii) a scale ,mobileobs obs icde in which mobileobsc determines the decay depending of the distance, ,obsid ,and (iii) a scale, 22()2 iie , based on the direction to the obstacle and with ,1a r c s i n ( )1 si o b s iDdensuring the generation of an attractor between two obstacles if the robot can pass through while ensuring the safety distance Ds. See 14 for more details. For ,mobile vobs if the expression adjusts the velocity towards ,obs obs ikd, but ensures that a minimum velocity ofminvis kept. To obtain the value of mobileobsfwe sum over all obstacles ,( ) ( )m o b i l em o b i l em o b i l e o b s iobsm o b i l e v m o b i l e vobs io b s o b s ifffff ( 17)C. Competitive Dynamics The weights for the competitive dynamics are controlled by equation (3) as explained above. Below are the competitive advantages and interactions for the two behaviors. 1) Target: The competitive advantage is set totar 0.5mobile whenever a target is present, otherwisetar 0.5mobile . The target behavior has the ability to interact with and suppress the obstacle avoidance behavior, when the ratio between the distance to the target and the closest object is sufficient to ensure the movement towards the target will be collision free. This is modeled as , m i n, l i m1 ( 1 t a n h ( ( ) ) )2 obsm o b i l e m o b i l e m o b i l et a r o b s g a i n i tt a rrrdd ( 18) 25 In which,minobsdis the distance to the closest obstacle, mobilegainris a gain constant giving how quickly the behavior should interact andlim 1mobileitr expresses the ratio between the distances to obstacle and target for which we will start to suppress obstacle avoidance. 2) Obstacle: The competitive advantage of the obstacle behavior is given by 00t a n h ( )m o b i l e m o b i l em o b i l eobs m o b i l e ( 19) In which mobile is the obstacle density as defined in Section III-D. The interaction is defined as 0,1 (1 t a n h ( ) ) (1 ) ) )2m o b i l e m o b i l eo b s t a r t a r o b s ( 20) The first part, 01 ( 1 t a n h ( ) )2 m o b i l e m o b i l e suppresses the target behavior when the obstacle density exceeds the threshold0mobile. The last part,1 mobiletar obs ,makes sure this only happens when the obstacle avoidance is not being suppressed due to,mobiletarobs . D. Calculation of Obstacle Density Given a set of distances, ,obsid , between the mobile platform and obstacles the density, , is calculated as ,1m a xi o b s id ( 21) This density function differs from 14 in which,obs idi e is used. The main problem with this formulation is that we cannot distinguish between many objects relative far away and a single object closed by. For example having 5 objects 2 meters away will give the same density as a single object 40 centimeters away. With the exponential function a single object in the scene can never cause to exceed 1. The threshold for switching to the obstacle avoidance behavior will thus have to be less than 1, but given a scene with multiple obstacles the threshold of 1 will often be too low. 26 Furthermore it is found that using,1obsidinstead of ,obsiemade tuning the parameters easier as we could now think of the density as the inverse of the distance. It also caused the density to grow very rapidly when getting close to an obstacle, thereby quickly forcing the behaviors to change. IV. CONTROL OF MANIPULATOR We will start by dividing the problem into two parts: 1) Determining the motion of the tool from the current position to the target while avoiding obstacles. 2) Inverse kinematics calculating joint velocities needed for the tool motion. The second part is a well understood problem, which in this work is solved using the inverse kinematics strategy presented in 23. This method incorporates both kinematics and dynamics limitations of the robot, such as joint position, velocity and acceleration limits. Furthermore this approach, based on quadratic optimization, has shown to be very robust with respect to singularities. The motion of the tool is controlled using the manipulators Target and Obstacle behaviors, to which the weights maniptarand manipobsare associated. As input the inverse kinematics needs a 6D velocity screw , , , , ,x y zx y z thus the behaviors must find a change , , , , ,m a n i px y zf x y z which can be integrated to give a desired tool velocity, des ,as .,( ( ) )m a n i p m a n i p m a n i p m a n i p m a n i pd e s c u r t a r t a r o b s o b s d i r o b s d i s td t f f f ( 22) Where maniptarf,.manipobsdirfand,manipobsdistfare the contributions from the target and obstacle avoidance behaviors. A. Target Behavior The inputs to the target behavior are the current and desired tool transformationscuranddesFrom these we can compute a desired 6D velocity-screw tar .To avoid requiring unrealistic fast motionstaris scaled such that m a x, ta rx y z v and m a x,x y zta rw w w wheremaxvandmaxhere represent the maximal 27 allowed linear and rotational velocities of the tool. Calculating ()m a n i p m a n i pt a r t a r c u r t a rf ( 23) We obtain a desired change to the current velocity. B. Obstacle-Behavior As input the obstacle avoidance behavior takes the current Cartesian velocity, ,v x y z and a set of closest obstacles as vectors, 3inR, giving direction and distance between tool and obstacle I . We now wish to compute a change to the Cartesian velocity based on the direction and distance to obstacles, denoted.manipobsdirf and ,manipobsdistfrespectively. 1) Dynamics for Direction: From the current velocity of the tool, v, and the vectorinwe compute the anglei, between the two as a r c s i n iiivnvn ( 24) The size of the change in direction of the tool is then calculated as 22, 22m a n i p io b s im a n i po b s i cnee ( 25) In which ,obsmanip is the strength of the repellor, manipobsccontrol the decay based the distance andcontrols the relation with the angle to the obstacle. is then used to calculate a desired change in the direction of the tool as iiivnvn ( 26) Summing up the contributions from all obstacles we can calculate the change in motion of the tool based on direction to obstacles as , ()m a n ipio b s d ir i vf ( 27) 2) Dynamics for Velocity: The dynamics of the velocity are controlled similar to Eq. (16). The contribution of obstacle I . is 28 ,m i n m i nm i n m a x ,m a x , m a x ,()0()m a n i p v m a n i p m a n i pobsm a n i pm a n i pio b s v e l im a n i p viiobsv f o rf o rv f o rf ( 28)Withm a x ,m a x , m a x ( , )m a n i po b s i ii kn Summing up over all obstacles the total contribution becomes ,m a n ip m a n ipo b s o b s v e l iiff ( 29) C. Competitive Dynamics 1) Target Behavior: As for the mobile platform the competitive advantage of the target behavior is set to 0.5 when a target is present and 0.5 otherwise. The competitive interaction of the target upon the obstacle behavior is again designed such that when the ratio between the distance to the target and to the nearest obstacle is greater then the thresholdlimmanipitrthe obstacle avoidance is suppressed. This is accomplished by l i m,m i n ( )1 ( 1 t a n h ( ( ) ) )2m a n i p iim a n i p m a n i pg a i n i tt o o lt a r o b st a rnrrd ( 30) In which tooltardis the distance between the tool and the target and mobilegainris a gain factor specifying how quickly to change the value of,maniptarobs . 2) Obstacle Behavior: The competitive advantage of the obstacle behavior is the same as in Section III-C, 00t a n h ( )m a n i p m a n i pm a n i p m a n i pobs ( 31) With the density calculated using Eq. (21), but with distances between obstacles and tool instead of obstacles and the mobile platform. The competitive interaction is defined as 0o b s , ,1 ( 1 t a n h ( ) ) ( 1 )2 m a n i p m a n i pm a n i p m a n i pt a r t a r o b s ( 32) In which the,(1 )maniptar obsterm helps to deactivate the obstacle avoidance as the tool gets close to the target. 29 D. Retract Behavior The retract behavior is operating directly in joint space. By defininghom e cu rq q q , where homeqis the home configuration to which it should retract, we can calculate the joint velocities as m a xm i n ,m a n i pr e t r a c t r e t r a c t m a n i pr e t r a c tqqqqq ( 33) Where maxq is the maximal velocity of the joints and manipretract is strength of the attractor. V. EXPERIMENTS The purpose of the experiments are primarily to demonstrate the coordination of the mobile platform and the manipulator. Previous work has already demonstrated the capabilities of the dynamical systems approach with respect to navigating a mobile robot through an environment 13 14 and guiding a manipulator around obstacles 16. The platform used in the experiments is shown in Figure 1 and consists of a Segway RMP200 and a Kuka Light Weight Robot equipped with a Schunk PG70 parallel gripper. The platform has a SICK LMS291 laser scanner for localization and obstacle avoidance and a tool mounted Unibrain Fire-I FireWire camera, used for aligning the gripper to the target. Unfortunately we did not have enough time to connect to the gripper and actually grasp the object. It thus only aligns and prepares to grasp, but never actually closes the gripper. As control framework we have used Microsoft Robotics Studio 1.5, which provided us with a tool for organizing the various inputs from sensors, outputs to actuators and ensuring concurrency of the different control algorithms. The motion of the Segway and the majority of the manipulator movement are based on a specified Cartesian location of the target. However, once the target is within view of the tool mounted camera, the guidance of the manipulator switches to rely on the visual input. Section V-A will explain details about the visual servoing approach, followed by Section V-B, which provides the test results. A. Visual Servoing For the final alignment of the gripper an eye-in-hand image based visual servoing approach is used. Feature extraction is done using the Simple Vision service in Microsoft Robotics Studio, which is able to identify colored blobs. In these experiments we are tracking a green 30 marker as illustrated in Figure 4. We wish the orientation of the tool to be fixed, thus only the 3 degrees of freedom (dof) associated with the position should be influence by the visual input. Two ofthese dof are controlled using the location of the blob, which should be centered in the image. The last dof is controlled by the size of the blob. (a) Move to target (time = 0s) (b) Visual servoing (time = 28s) (c) Move to place position (time = 40s) (d) Place item (time = 72s) Fig . 3 . Experiment with the mobile manipulator. Weights the top right corner to weights corresponding to the given situation. Fig. 4. Feature detection using Microsoft Robotics Studios SimpleVision service. The black and white 31 border marks the feature identified. B. Test Results The task of the mobile manipulator is, as illustrated in Figure 3, to move a bottle from the table in the middle of the image to the box located to the far right. The weights associated with the Mobile, Manipulator Retract and Manipulator Target behaviors can be found in Figure 5. Initially both the Manipulator Retract and the Mobile behaviors are active causing the platform to move towards the target, while keeping the arm in its home configuration. After about 7 seconds the object gets within reach, thus the Manipulator Retract behavior deactivates and Manipulator Acquisition is activated. Shortly after the Segway behavior is also deactivated, to let the manipulator pick up the object without disturbances. However, the motion of the manipulator causes the Segway to drift, thus after a little while the mobile platform is reactivated until about time equals 20s, where it has again reached the desired position relative to the target. The visual servoing then aligns the gripper to the bottle as illustrated in Figure 3(b). After around 30 seconds the bottle should have been picked up by the gripper and a new target is given, causing the Manipulator Retract behavior to reactivate and Manipulator Acquisition to deactivate. At this point the mobile platform behavior is also activated, but is quickly suppressed while the arm is being retracted. Afterwards the control moves the platform to the desired location where the manipulator is activated to place the object. Fig. 5. Weights of the behaviors while operating VI. CONCLUSION In this paper it has been presented how the dynamic systems approach can be applied to 32 mobile manipulation. The contributions include a two level approach in which competitive dynamics are used both for the overall coordination of a mobile platform and a manipulator as well as the obstacle avoidance and target acquisition behaviors. The approach has been verified first

温馨提示

  • 1. 本站所有资源如无特殊说明,都需要本地电脑安装OFFICE2007和PDF阅读器。图纸软件为CAD,CAXA,PROE,UG,SolidWorks等.压缩文件请下载最新的WinRAR软件解压。
  • 2. 本站的文档不包含任何第三方提供的附件图纸等,如果需要附件,请联系上传者。文件的所有权益归上传用户所有。
  • 3. 本站RAR压缩包中若带图纸,网页内容里面会有图纸预览,若没有图纸预览就没有图纸。
  • 4. 未经权益所有人同意不得将文件中的内容挪作商业或盈利用途。
  • 5. 人人文库网仅提供信息存储空间,仅对用户上传内容的表现方式做保护处理,对用户上传分享的文档内容本身不做任何修改或编辑,并不能对任何下载内容负责。
  • 6. 下载文件中如有侵权或不适当内容,请与我们联系,我们立即纠正。
  • 7. 本站不保证下载资源的准确性、安全性和完整性, 同时也不承担用户因使用这些下载资源对自己和他人造成任何形式的伤害或损失。

评论

0/150

提交评论