【作者】 姜莹；

【导师】 王硕玉； 白保东；

【作者基本信息】 沈阳工业大学 ， 医学电磁工程， 2013， 博士

【摘要】 全球老龄化问题日益严峻，各种疾病、运动损伤及交通事故所引发的下肢运动功能障碍患者显著增加，这类患者除了早期的手术治疗和必要的药物治疗外，正确、科学的康复训练对于患者运动功能恢复将起到重要作用。本文以辽宁省教育厅项目和辽宁省自然科学基金项目为依托，对全方位移动下肢康复机器人的运动学、动力学建模，康复训练过程中机器人发生重心偏移时的轨迹跟踪控制进行了深入研究。该项研究目前在国内正处于起步阶段，具有广阔的发展空间，将为全方位移动下肢康复机器人系统早日应用奠定理论及技术基础。全方位移动下肢康复机器人是一强耦合、非线性且含有诸多不确定因素的复杂系统。执行机构的机械误差、自身质量和转动惯量、环境、轮与地面的打滑和摩擦、机械震动等都会对机器人运动特性产生影响。尤其，受负载因素影响全方位移动机器人在辅助患者进行康复训练时重心位置是不确定的，重心偏移时机器人的运动控制已成为此类康复机器人亟待解决的一个重要问题。论文主要研究工作包括以下四个方面：（1）全方位移动下肢康复机器人考虑重心偏移时数学模型的建立。结合全方位轮式移动下肢康复机器人的机构特点进行运动学和动力学分析，分别讨论康复机器人理想空载情况、考虑荷重变化但未产生重心偏移以及考虑荷重变化并产生重心偏移时全方位移动下肢康复机器人运动学模型和动力学模型的建立。（2）针对全方位移动下肢康复机器人提出一种目标轨道加速度、速度信息有效利用的新型控制器设计。为设计该控制器，改进描述全方位移动下肢康复机器人重心偏移时动力学模型，利用系统模型中物理矩阵之间的关系等式，提出一种全新的控制器设计策略。控制器中有效包含目标轨道的加速度、速度信息，可以实现快速、准确地跟踪效果。并在满足患者质量定常、重心偏移定常可知条件下，基于Lyapunov定理证明了此控制律的稳定性。对全方位移动下肢康复机器人所提出的这种将目标轨道加速度、速度信息有效融入控制器设计的方法，通过与常规PD控制器设计方法对比进行了深入的仿真研究，验证了设计方法的有效性。（3）针对全方位移动下肢康复机器人重心未知、时变偏移提出一种鲁棒补偿控制策略。考虑实际康复机器人训练过程中重心往往是时变且未知的，进一步提出一种有针对性解决重心偏移模型中含有时变未知项的鲁棒控制策略，通过设计一个具有H∞跟踪性能的鲁棒补偿控制器来补偿系统中的未知时变部分，既有效地改善了康复机器人重心时变偏移时的运动性能，同时又使系统具有良好的抗干扰能力。另外，该控制器设计基于全方位移动下肢康复机器人四轮控制力的限定约束关系，对系统重心偏移动力学模型进行改进描述，打破了以往全方位移动下肢康复机器人控制器设计严重依赖系统模型中的重心时变项和控制力的不可独立性，为全方位移动下肢康复机器人重心偏移时控制器设计提供了一个新的解决问题思路。（4）全方位移动下肢康复机器人实验平台及轨迹跟踪控制实验。基于日本高知工科大学“步行王”全方位移动下肢康复机器人样机，摄像机和上位计算机组成的实验平台，对本文所提出的有效利用全方位移动下肢康复机器人目标轨道加速度、速度信息的控制方法和有效针对重心未知、时变偏移所提出的鲁棒补偿控制方法分别进行了直线轨迹跟踪控制实验研究，实验验证了本文所提出的运动控制策略具有可行性。全方位移动下肢康复机器人的研制及其运动控制算法的研究在国内尚起步，算法的有效性对康复机器人的硬件、联机、可靠性、实时性等方面均有极高要求。本论文研究旨在针对重心偏移时控制器设计改善康复运动的训练效果，为我国下肢康复机器人领域的发展起到微薄推动作用。更多还原

【Abstract】 The global population aging is increasingly serious and patients with lower-limbdysfunction caused by various diseases, injuries and traffic accidents have increasedsignificantly. Correct, scientific rehabilitation for these patients plays an important role, inaddition to early treatments and necessary drug treatments. In this dissertation, the kinematic,dynamic modeling of an omnidirectional lower-limb rehabilitation robot and the trajectorytracking control under the centre-of-gravity shift during walking training are studied based onthe research projects of Liaoning Provincial Department of Education and Liaoning ProvincialNatural Science Foundation. This research, which our country is short of, so has a broad outlookand can provide theoretical and technical base for the application of the omnidirectional lower-limb rehabilitation robot.The omnidirectional lower-limb rehabilitation robot is a strongly coupled, complexnonlinear system including many uncertainties. The robot’s motion is affected by mechanicalerrors of actuators, mass and inertia, environments, the sliding and friction between the wheeland ground and mechanical vibration. Especially, the center of gravity of the omnidirectionalmobile robot during rehabilitation affected by load factors is uncertain. The motion control todeal with the centre-of-gravity shift has become an important problem in such field. Thisdissertation mainly includes the following aspects:(1) Mathematical modeling of omnidirectional lower-limb rehabilitation robot with regardfor centre-of-gravity shift. Combined with the analyzing of the kinematics and dynamicscharacteristics of a four wheeled omnidirectional robot, the models of the robot kinematics anddynamics are established, respectively the ideal case with no load, with consideration of themass of the load, but not consideration of the centre-of-gravity shift and with regard for thecentre-of-gravity shift.(2) For the omnidirectional lower-limb rehabilitation robot, a new controller by using theacceleration and velocity information effectively of the target trajectory is proposed. This new controller is proposed by using the relation equations of certain physical matrices of the systemdynamic model perfectly. By including the acceleration and velocity information, the speed ofcvergence and accuracy of the tracking error are well improved. The stability of this controllerwas proved by Lyapunov theorem under the condition of the constant mass of the patient. Thecontroller was simulated to verify the correctness by comparison with PD controller.(3) A robust control strategy under the unknown and time-varying of the centre-of-gravityis proposed. Considering that the centre-of-gravity of the robot is always unknown and time-varying, a robust compensation control strategy aimed to solve the time-varying and unknownof the centre-of-gravity shift is developed. By developing an adaptive robust controller withH∞tracking performance to eliminate the time-varying and unkown part of the system, themotion performance and the interference suppression of the rehabilitation robot has both beenimproved. Based on specifying constraint relationship on four wheels’ control force, the newstrategy broke the previous limitation and dependence of the controller to the time-varyingmatrix of the system model.(4) Experimental platform and research on trajectory tracking control of theomnidirectional lower-limb rehabilitation robot. Based on the platform consisting of the robotdesigned by Kochi University of Technology, the camera and the computer, the feasibilities ofthe above proposed trajectory tracking control algorithms of chapter3and chapter4arevalidated through tracking basic line trajectory.The research on the omnidirectional lower-limb rehabilitation robot and its motion controlalgorithms are yet to start in our country. The effectiveness of the algorithms is also closelyrelated to the hardware of robot, online, reliability and real-time. The purpose of this dissertationis to improve the effect of rehabilitation based on control methods during the centre-of-gravityshift and to provide modest promotion to the development of lower-limb rehabilitation robot.更多还原