文章摘要
刘磊,杨鹏,刘作军,宋寅卯.采用改进D-S证据融合的动力型假肢运动模式预识别[J].高技术通讯(中文),2021,31(3):298~305
采用改进D-S证据融合的动力型假肢运动模式预识别
Walking mode pre-recognition of lower limb prosthesis based on the data fusion of improved D-S theory
  
DOI:10.3772/j.issn.1002-0470.2021.03.010
中文关键词: 假肢关节; 传感器; 改进D-S(Dempster/Shafer)证据理论; 运动模式预识别
英文关键词: prosthetic joint, sensor, improved D-S (Dempster/Shafer) theory, walking mode pre-recognition
基金项目:
作者单位
刘磊  
杨鹏  
刘作军  
宋寅卯  
摘要点击次数: 1733
全文下载次数: 1088
中文摘要:
      动力型假肢是目前康复工程研究的热点问题之一,动力型假肢穿戴者行走的各路况运动模式的预识别是关键的一个环节。为了能够充分地描述路况信息,本文采集髋关节角度、加速度运动信息和前后脚底压力信息作为识别信号源,采用改进的D-S(Dempster/Shafer)证据理论方法将陀螺仪传感器和加速度传感器信息融合,从而对假肢穿戴者行走的各路况运动模式进行有效的预识别。实验结果表明,基于改进证据理论数据融合的动力型下肢假肢运动模式预识别方法可有效地对上楼梯、下楼梯、上斜坡、下斜坡、平地行走5类运动模式进行预识别,预识别率高达97.6%。该技术对动力型假肢的研制具有重要的实用价值。
英文摘要:
      In order to improve the uncoordinated control of prosthesis caused by the delay effect of dynamic prosthetic motion decision program and prosthetic actuator, a walking mode pre-recognition method based on the data fusion of improved D-S (Dempster/Shafer) theory is used to drive prosthetic joint motion in advance and real-time adjust prosthetic gait, which can achieve the best coordinated control effect. The prosthetic motion information of gyroscope and acceleration sensors in the prosthetic socket are used as the data source, and the foot pressure switch information auxiliary analysis is used. The improved method of D-S theory combined with distance function is used to fuse the correlation coefficient, thereby to pre-recognize walking mode effectively. The experimental results show that five walking mode including upstairs, downstairs, uphill, downhill, level walking can be pre-identified by walking mode pre-recognition of power lower limb prosthesis based on the data fusion of improved D-S theory. The method has important practical value to develop the technology of dynamic prosthesis.
查看全文   查看/发表评论  下载PDF阅读器
关闭

分享按钮