下肢康复机器人运动控制策略的研究

发布时间：2018-07-04 10:44

本文选题：下肢康复机器人 + 外骨骼　；参考：《哈尔滨工业大学》2017年硕士论文

【摘要】：康复机器人是医疗机器人的重要分支,涉及康复医学、机器人学及控制学等多个领域,致力于瘫痪患者的肢体康复。近年来,随着老龄化现象的日益严重,肢体瘫痪的患者也逐渐增多。本文针对下肢瘫痪的患者,设计了一款坐卧式下肢康复机器人,针对其控制策略进行了研究,包括患者使用其进行被动运动和主动运动时进行的控制,并通过相应的仿真和实验进行验证。为了更好地控制机器人的运动,需要详细的了解本课题的实验平台。本文研究的坐卧式下肢康复机器人具有3个自由度,可以满足患者日常的步态训练。利用拉格朗日方程对机器人系统进行动力学分析,并推导出电动缸的驱动力与力矩的关系。通过动力学方程得到关节驱动力矩,然后利用ADAMS软件建立康复机器人系统的动力学模型,通过仿真实验得到驱动力矩,两者比较验证了动力学方程和ADAMS模型的正确性。当患肢完全瘫痪时,需要完全由外力带动运动,即被动运动训练,而被动运动控制可看作是轨迹跟踪控制。本课题提出的被动运动控制策略结合了三种控制方法:基于计算力矩法的PD反馈控制主要用于系统标称模型的控制,RBF神经网络用于逼近补偿系统的模型的不确定项,自适应鲁棒控制用于补偿RBF神经网络的逼近误差和外界干扰。本文运用李雅普诺夫稳定性理论验证了其稳定性,并通过仿真实验验证算法的有效性。当患肢逐渐康复,有一定的运动能力时可以进行主动运动训练,本文通过基于位置的阻抗控制对患肢进行期望力的跟踪。本文首先对人机接触阻抗进行了研究,然后分析了阻抗参数对系统控制性能的影响,利用粒子群优化算法优化阻抗参数,最后利用参考模型自适应控制提高系统的鲁棒性,并通过仿真实验验证了算法的有效性。完成对坐卧式下肢康复机器人实验系统的搭建,包括硬件系统和软件系统。针对实验平台的特点,对电机和驱动器进行了相应的设置,并对传感器进行了标定,使其可以实时的检测外界环境。为了保证系统的运转精度和安全性,对系统相应的输入和输出进行了详细推导,并设计了软件限位。最后通过实验验证了本文提出的基于计算力矩法的神经网络鲁棒控制和基于MRAC的自适应阻抗控制的在提高控制精度方面的有效性。
[Abstract]:Rehabilitation robot is an important branch of medical robot. It is involved in rehabilitation medicine, robotics and control. In recent years, with the increasing aging phenomenon, the number of patients with limb paralysis is increasing. In this paper, a sitting and horizontal lower limb rehabilitation robot is designed for patients with lower extremity paralysis. The control strategy is studied, including the passive and active motion control. And through the corresponding simulation and experiment to verify. In order to better control the motion of the robot, we need to understand the experimental platform in detail. The robot has 3 degrees of freedom, which can meet the patients' daily gait training. The dynamic analysis of robot system is carried out by using Lagrange equation, and the relation between driving force and torque of electric cylinder is deduced. Then the dynamic model of rehabilitation robot system is established by using Adams software, and the driving torque is obtained by simulation experiment. The correctness of the dynamic equation and Adams model is verified by comparing the two equations. When the affected limb is completely paralyzed, it needs to be driven by external force, that is, passive motion training, and passive motion control can be regarded as trajectory tracking control. The passive motion control strategy proposed in this paper combines three control methods: PD feedback control based on calculating torque method is mainly used to control the nominal model of the system and RBF neural network is used to approximate the uncertainty of the model of compensation system. Adaptive robust control is used to compensate the approximation error and external disturbance of RBF neural network. In this paper, Lyapunov stability theory is used to verify its stability, and the effectiveness of the algorithm is verified by simulation experiments. When the affected limb recovers gradually and has certain movement ability, the active movement training can be carried out. In this paper, the expected force of the affected limb is tracked by the impedance control based on position. In this paper, the impact of impedance parameters on the control performance of the system is analyzed, and the particle swarm optimization algorithm is used to optimize the impedance parameters. Finally, the reference model adaptive control is used to improve the robustness of the system. The effectiveness of the algorithm is verified by simulation experiments. The experiment system of sitting and horizontal lower limb rehabilitation robot is built, including hardware system and software system. According to the characteristics of the experimental platform, the motor and driver are set up, and the sensor is calibrated so that it can detect the external environment in real time. In order to ensure the operation precision and security of the system, the corresponding input and output of the system are deduced in detail, and the software limit is designed. Finally, the effectiveness of the proposed neural network robust control based on the computational moment method and the adaptive impedance control based on MRAC in improving the control accuracy is verified by experiments.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP242

【参考文献】