下肢假肢穿戴者跑动步态识别与膝关节控制策略研究  

Research on running gait recognition and motion control strategy for lower limb prosthesis wearer

作　　者：刘作军[1] 高新智 赵晓东[1] 陈玲玲[1] Liu Zuojun;Gao Xinzhi;Zhao Xiaodong;Chen Lingling(Engineering Research Center of Intelligent Rehabilitation and Detecting Technology,Hebei University of Technology,Tianjin 300130,China)

机构地区：[1]河北工业大学智能康复装置与检测技术教育部工程研究中心

出　　处：《仪器仪表学报》2018年第7期74-82,共9页Chinese Journal of Scientific Instrument

基　　金：国家自然科学基金(F030604,61703135);河北省自然科学基金(F2018202279)项目资助

摘　　要：为研究假肢穿戴者跑动状态以及假肢膝关节的跑动控制,首先通过人体多源运动信息采集,获取髋、膝关节加速度信号、大腿股直肌部位肌电信号和足底压力信号,使用切比雪夫I型滤波方法进行降噪并提取信号特征。利用相关性分析方法,结合足底压力信号实现对跑动状态识别。在此基础上,利用粒子群优化的支持向量机方法进行信号分析,实现对6个跑动相位的识别。以四连杆假肢为被控对象,以人体跑动信号的特征点为各相位的转移条件,提出了基于有限状态机的假肢跑动控制策略。通过多项式拟合,建立了假肢膝关节控制模型,并进行了下肢假肢测试平台的验证。测试结果表明,所提方法可以实现跑动步态与相位的有效识别,并控制假肢膝关节做出正确的跑动动作。The running gait recognition of the prosthesis wearer and the control strategy of the prosthetic knee joint are studied in this paper. Firstly,a human multi-source motion information system is used to collect the acceleration signals of the hip and knee joint,the electromyography signals of the rectus femoris,and the plantar pressure signal. The signals are denoised and extracted by the Chebyshev type I filtering method. The correlation analysis is combined with the plantar pressure signals to realize the recognition of the running state. Moreover,the particle swarm optimization and support vector machine are utilized to identify the six phases in the running process. A four-link prosthesis control strategy for running motion based on finite state machine（ FSM） is proposed. The key switching points in motion signals are selected as the state transition conditions. The control model is established by polynomial fitting and is tested on the lower limb prosthesis test platform. The experiments on the lower limb prosthesis test platform show that the proposed method can achieve effective recognition and control of prosthetic knee joint in running motion.

关 键 词：假肢 步态识别 支持向量机 有限状态机 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置] TH122[自动化与计算机技术—控制科学与工程]