康复机器人上肢的模糊PID和神经网络阻抗控制  

Upper Limb Control Strategy of Rehabilitation Robot Based on Fuzzy PID and Neural Network Impedance Control

作　　者：穆海芳[1] 郭凯[1,2] 胡波 MU Haifang;GUO Kai;HU Bo(School of Mechanical and Electronic Engineering,Suzhou University,Suzhou 234000,China;Anhui Provincial Engineering Laboratory on Information Fusion and Control of Intelligent Robot,Wuhu 241002,China)

机构地区：[1]宿州学院机械与电子工程学院,安徽宿州234000 [2]安徽省智能机器人信息融合与控制工程实验室,安徽芜湖241002

出　　处：《济宁学院学报》2022年第2期89-92,97,共5页Journal of Jining University

基　　金：宿州学院科研平台项目“康复机器人上肢运动控制策略研究”(2020ykf10)。

摘　　要：肢体运动功能障碍患者借助康复机器人进行辅助康复训练时,为了提高康复效果,需要选择合适的控制策略。目前,阻抗控制是一种广泛使用的力控制方法,但传统阻抗控制方法不能够根据患肢病情的变化情况及时调整阻抗参数。在传统阻抗控制策略基础上,提出一种基于模糊PID结合神经网络阻抗控制的方法,建立了系统总体结构、模糊PID控制器结构以及神经网络阻抗控制器结构,采用模糊PID算法实现机器人的位置控制,神经网络阻抗控制实时调整阻抗控制参数。取某两连杆机械臂做仿真实验,结果表明,与传统的阻抗控制方法比较,该方法控制的系统具有更好的跟踪性和协调性。When patients with limb motor dysfunction using rehabilitation robot doing assisted rehabilitation training, in order to improve the rehabilitation effect, they need to choose the right control strategy. Currently, impedance control is a widely used force control method, but traditional impedance control methods cannot timely adjust impedance parameters according to the change of limb condition. So based on the traditional impedance control strategy, this paper presents an Impedance Control Method Based on Fuzzy PID and Neural Network, establishes the overall structure of the system, structure of fuzzy PID controller and Neural Network. Using Fuzzy PID Algorithm realizing Position Control of Robot and Neural network impedance control real-time adjustment of impedance control parameters. A two-link manipulator simulation experiment shows that, compared with traditional impedance control methods, the system controlled by this method has better traceability and coordination.

关 键 词：康复机器人 阻抗控制 模糊神经网络 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]