Design and locomotion analysis of a novel deformable mobile robot with two spatial reconfigurable platforms and three kinematic chains

Abstract

A novel five degrees of freedom deformable mobile robot composed of two spatial reconfigurable platforms and three revolute–prismatic–spherical kinematic chains acting in parallel to link the two platforms is proposed to realize large deformation capabilities and multiple locomotion modes. Each platform is an improved deployable single degrees of freedom three-plane-symmetric Bricard linkage. By taking advantage of locomotion collaborating among platforms and kinematic chains, the mobile robot can fold into stick-like shape and possess omnidirectional rolling and worm-like motions. The mechanism design, kinematics, and locomotion feasibility are the main focus. Through kinematics and gait planning, the robot is analyzed to have the capabilities of rolling and turning. Based on its deformation, the worm-like motion performs the ability to overcome narrow passages (such as pipes, holes, gaps, etc.) with large range of variable size. Dynamic simulations with detailed three-dimensional model are carried out to verify the gait planning and provide the variations of essential motion and dynamic parameters in each mode. An experimental robotic system with servo and pneumatic actuation systems is built, experiments are carried out to verify the validity of the theoretical analysis and the feasibility of the different locomotion functions, and its motion performances are compared and analyzed with collected data.

Reconfigurable/Foldable Overconstrained Mechanism and Its Application

“…For this linkage, the Denavit and Hartenberg notations are as follows: one kinematical variable θ i -the rotation angle of the joint, and three geometrical parameters: ai the bar lengths, α i the twist angle between two successive joints (i) and (i + 1), and di the offset distance between two elements (i − 1) and (i). Other applications of overconstrained mechanisms, foldable/deployable network based on Bennett linkage, were presented by Gan and Pellegrino [14], Baker [15] and Chen [16], as well as more recently by Ding [17] and Song [18].…”

“…Other applications of overconstrained mechanisms, foldable/deployable network based on Bennett linkage, were presented by Gan and Pellegrino [14], Baker [15] and Chen [16], as well as more recently by Ding [17] and Song [18]. This paper is focused on research results of another type of foldable/reconfigurable structure, starting from a 6R symmetric mechanism, in a particular spatial disposition and with good estimated application in the automotive industry.…”

State of the Art of Multi-mode Mechanisms

“…雅可比矩阵与机构的位形密切相关， 可以描述 机构末端运动速度与关节运动速度之间的映射关 系 [95] 。在得知输入或输出速度的情况下，通过雅 可比矩阵便可以求得机构动平台的瞬轴面。CHEN 等 [96] 利用雅可比矩阵法分别给出了 3-RPS 角台并 联机构的 2 种瞬轴面，用以展示该机构的卡当运动 特性和常规运动特性。 4 多模式机构的应用研究 机构研究的最终目标是为了投入实际应用，满 足现代化生产作业的需求。然而，多模式机构的发 展仍处于初级阶段，距离全面实现工程应用还有诸 多问题有待解决。目前，关于多模式机构应用方面 的研究主要集中于医疗康复、航空航天和机械加工 等领域。 在医疗康复领域，HE 等 [97] 基于多模式机构设 计了一种可实现自由度变化的腹腔微创手术机器 人；该机器人通过运动模式转换可提高末端执行器 的刚度和输出力以完成不同的手术任务。 TSENG 等 [98] 利用多模式连杆机构提出了一种康复训练设备；依 靠不同的运动模式该设备可分别完成臀部和小腿部 位的康复训练。依赖具有两种运动模式的 3-RPS 机 构，NURAHNI 等 [99] 设计了一种新型踝关节康复设 备；通过运动模式的切换，该设备可以完成不同的 康复训练动作。 在航空航天领域，BAIGUNCHEKOV 等 [100] 提 出了一种多模式定位并联机构； 在不同运动模式下， 该机构可以完成不同的定位任务。由于轻便、可折 叠等优势， NASA 于 2012 年已着手研究基于连杆的 星球探测机器人； 一款名为 SuperBall 的星球探测机 器人可通过机构的整体变形实现滚动运动，从而实 现在复杂地形中的移动 [101] 。受此启发，北京交通大 学的姚燕安教授团队 [102][103][104][105] 开发了一系列可用于复 杂地形探测的多模式移动机器人；利用连杆机构的 多种运动模式，此类机器人可实现各式各样的移动 运动形式(滚动、步行和蠕动等)。 在机械加工领域，为了满足灵活、高效的加工 需求，AZULAY 等 [106] 设计了一种可以在较快进给 速率下保持高刚度的变自由度铣床，其能够针对不 同工件完成相应的细铣任务。基于 3-CPU 机构， CARBONARI 等 [70] 设计了一种可以完成三维转动 和三维移动的加工机床。 [107] 。另外，在走向应用 的道路上，以多模式机构运动学和动力学分析为基 础的性能评价和尺度综合研究也有待进一步开展。 参 考 文 献 [7] ZHANG K T，FANG Y F，WEI G，et al Advances in reconfigurable mechanisms and robots [M]. London ： Springer，2012.…”

Adaptive decentralized fuzzy compensation control for large optical mirror processing systems

“…Owing to the coupling, mutual disturbance and external environment interference among multiple heterogeneous robots, the processing accuracy cannot meet the system requirements. Moreover, the control error generated by multiple robots accumulates the error of the final optical mirror (Ding et al , 2017; Jiang et al , 2018). This leads to an uncertainty in the dwell time and causes some local areas to be over- or underprocessed, thereby prolonging the entire processing cycle and even leading to the mirror being discarded (Li et al , 2020).…”