【作者】 王企远；

【导师】 钱晋武；

【作者基本信息】 上海大学 ， 机械电子工程， 2011， 博士

【摘要】 步行康复训练机器人系统是一种对下肢功能障碍患者进行步行康复训练的现代医疗设备。使用机器人对下肢功能障碍患者进行步行康复训练,是当前世界公认的先进的康复医疗手段。近些年来,我国患下肢运动障碍的病人不断增多,社会对步行康复治疗的需求也越来越大。然而,由于我国目前尚无成熟的步行康复训练机器人产品,国外的此类产品价格又很高,所以国内医疗机构还没有真正运用这项医疗技术。加快自主研发这种设备,对提升我国医疗水平、提高患者生活质量、减轻社会负担都具有重要的意义。本文所研究的助行腿的步态规划与运动控制是步行康复训练机器人系统研究中的一部分,也是其中的关键技术。本文所做的研究是制定步行康复训练机器人助行腿的步态规划,建立操纵助行腿运动的控制系统,主要研究内容包含以下几方面。(1)研究了步态特征参数的测试原理与计算方法。建立了由运动捕捉系统和跑步机组成的步态特征参数测试平台,研究了运动捕捉系统中局部坐标系的建立方法、坐标变换方法以及关节角度的计算方法,并通过实验检验了它们的可行性与正确性。在此基础上,确定了步态特征参数的具体测试方法与操作步骤。(2)研究了人体下肢髋关节、膝关节和踝关节转角变化规律的数学建模方法。采用非线性曲线拟合的方法分别建立了髋关节、膝关节和踝关节转角变化规律的数学模型。对数学模型进行了残差分析,结果表明,构建出的数学模型能够如实反映出人体行走时的步态特征,能满足助行腿步态规划的需要。(3)针对助行腿的工作任务和工作环境,制定了一套完整的步态规划方法。根据人体下肢的行走特征,将助行腿的训练过程划分为起步、周期步和止步三个阶段,确定了每个阶段的功能和作用。根据助行腿在各阶段的运动特征与任务要求,分别制定出了每一阶段的步态规划。最后将三个阶段的步态规划按顺序组合在一起,合成为一套适用于助行腿步行训练的完整步态规划。(4)研究了助行腿的运动控制策略和运动控制数据生成方法,并应用于助行腿的运动控制系统中。根据伺服电机的工作原理和助行腿的结构特点,确定了运动控制系统所用的控制策略,建立了基于关节转角位置的闭环控制方法。根据助行腿的关节结构和驱动器的动力传递顺序,求出了电机轴与对应关节轴之间的转角传递函数,并以此为基础编制了运动控制数据生成程序,将规划出的目标运动轨迹转换为电机驱动器所需的脉冲控制指令数据。(5)通过实例对所研究的步态规划方法与运动控制系统进行了验证。用测试出的一组周期步关节角变化轨迹数据,规划出助行腿各关节的目标运动轨迹,按规划出的目标运动轨迹计算出助行腿各关节的运动控制指令数据,在步行康复训练机器人系统样机上进行了模拟训练试验。试验结果表明,助行腿能够按照规划的步态带动人体模型进行训练,助行腿的行走步态相对于规划步态的误差很小,能够达到以人体正常步态带动患者训练的目的。更多还原

【Abstract】 Gait Training Robot (GTR) is a kind of modern medical equipment for the rehabilitation training of individuals with locomotor disfunction of the lower limbs. Currently, performing the rehabilitation training with GTR has become a world recognized rehabilitation method with obvious treatment effects. In recent years, with the increase of individuals with locomotor disfunction of the lower limbs in China, the needs for the gait rehabilitation training have become more and more. Because the prices of the overseas GTR products are very high and there isn’t home-made GTR product with mature technology, this medical equipment has not been used in domestic medical establishments. So development this kind equipment has great significance to improve domestic medical treatment level and to improve the patients’life quality.The dissertation is part of the study on the key techniques of Gait Training Robot System. The main works of this dissertation are establishing the gait planning method and motion control system for the Powered Gait Orthosis (PGO) of the GTR System. The main contents of this dissertation are as follows:(1) Testing and computing methods of the gait characteristic parameters. A gait characteristic parameter test bench is built, which consists of a motion capture system and a treadmill. The local coordinate system creation method, coordinate transformation method and the computing methods for the parameters of joint angles are studied. The feasibility and correctness of these methods have been tested by experiments.(2) Mathematical model creation method for the hip joint, knee joint and ankle joint angle data. The hip joint, knee joint and ankle joint angle data mathematical model are established by using non-linear curve-fitting methods and the residual analyses are made. The residual analyses results demonstrate that these created mathematical models can describe the human gait characteristics exactly and meet the demands of human natural gait planning for the PGO.(3) A set of overall gait planning methods for the PGO are established according to the PGO’s work tasks and work environments. The overall walk training process consists of three gait phases: start gait, cycle gait and end gait. The gait of each gait phase is planed according to its motion characteristic and task demand. In the end, a set of overall gait planning for the PGO are established by combining the three gait phases in a certain sequence.(4) The motion control strategy and its data generation method for the PGO are studied and applied to the motion control system of the PGO. According to the operating principle of servo motor and structural features of PGO, the motion control strategy is confirmed and the closed loop position control method, which based on the joint angle data, is established. According to the joint structure and motion transmission order of the joint actuator, the motion transfer functions from motor axis to joint axis are derived. Based on the motion transfer functions, the motion control data generation program is programmed, which can transform the target motion trajectories to pulse instructions needed by the servo motor drivers.(5) The gait planning methods and motion control system for the PGO have been tested through experiments. First detect and calculate a set of joint angle data of cyclic gait, then plane the target motion trajectories of the three joints, and then transform the target motion trajectories to pulse instructions data. At last, the gait training experiments are carried out on the GTR prototype with a dummy. The experimental results show that the PGO is able to guide the dummy’s legs to move in the planned gait pattern on treadmill and the motion trajectory errors are very small. The goal of driving the patient’s legs to move in human natural gait pattern can be achieved by using the GTR prototype.更多还原