基于步态数据的下肢康复机器人控制设计

Control Design of Lower Limb Rehabilitation Robot Based on Gait Data

为了使下肢康复机器人的辅助康复训练更加符合人体运动特性,为下肢运动功能障碍的患者提供更加安全、有效的康复训练,提出一种基于人体步态数据和径向基函数(RBF)神经网络的下肢康复机器人控制策略。首先,以三维动作捕捉系统获取的健康人体步态数据为系统的期望输入;其次,通过RBF神经网络自适应控制器产生驱动机器人关节运动的力矩,使机器人的运动轨迹跟踪期望轨迹;然后,采用基于前馈控制的校正模型对机器人的输入力矩进行补偿校正,达到实时校正的目的,从而实现更好的跟踪效果;最后,通过基于实验数据的仿真结果,验证了该方法的可行性和有效性。

In order to make the assisted rehabilitation training of the lower limb rehabilitation robot(LLRR)more consistent with human motor characteristics, and to provide more safe and effective rehabilitation training for patients with lower limb motor dysfunction, a control strategy of the LLRR based on human gait data and radial basis function(RBF) neural network is proposed. Firstly, the gait data of healthy human body acquired by 3 D motion capture system is taken as the expected input of the system. Secondly, the RBF neural network adaptive controller generates the torque to drive the joint motion of the LLRR, so that the motion trajectory of the robot can track the desired trajectory. Then, the modified model based on feedforward control is used to compensate and correct the input torque of the robot, so as to achieve real-time correction and better tracking effect. Finally, the simulation results based on experimental data demonstrate the feasibility and effectiveness of the proposed method.