摘要
下肢外骨骼需要通过识别穿戴者的运动意图为穿戴者日常活动提供助力,然而当前的研究很少关注能够提供新受试者意图信息的下肢运动模式预测.为此,本文提出了一种基于多传感器信息融合和迁移学习的下肢运动模式预测方法.本文首先设计了一个下肢运动模式预测模型,采用长短时记忆单元(Long-Short Term Memory,LSTM)提取表面肌电信号(surface ElectroMyoGraphy,sEMG)中的模式特征,然后将sEMG的模式特征与关节角度特征融合预测下肢运动模式.考虑到受试者之间的生理信号差异,本文设计的迁移学习策略分两步训练预测模型,第一步在源域受试者数据集上预训练模型,第二步冻结sEMG模式特征提取器的网络权值,并在目标域数据集上微调全连接层.实验采集了受试者自由行走和穿戴外骨骼行走的数据.通过预测时间长度为100 ms的实验可以得出,所提出的方法分别能够有效提升新受试者自由行走状态下和穿戴外骨骼行走时9.53%和8.29%的运动模式预测准确率.实验结果表明,所提出方法可通过提升新受试者运动模式预测准确率,从而保障下肢外骨骼可靠的人体运动意图感知.
Lower limb exoskeletons require the capability to identify the user’s lower-limb motion intentions to pro⁃vide support during daily activities.However,existing research rarely focuses on predicting locomotion modes that provide user intention for new subjects.To bridge this gap,this study proposes a novel method for lower-limb locomotion mode pre⁃diction based on multi-sensor signal fusion and transfer learning.The study first designs a prediction model that utilizes long-short term memory(LSTM)networks to extract pattern features from surface electromyography(sEMG)signals.These sEMG features are then fused with joint angle features to predict lower-limb locomotion modes.Considering the in⁃ter-subject variability in physiological signals,the method employs a two-step training process using transfer learning.First,the model is pre-trained on a source domain dataset.Next,the weights of the sEMG feature extractor are frozen,and the ful⁃ly connected layers are fine-tuned using a target domain dataset.Experimental data are collected from subjects performing both normal walking and exoskeleton-wearing walking.Experimental results with a prediction time of 100 ms demonstrate that the proposed method enhances motion pattern prediction accuracy by 9.53%during free walking and by 8.29%during exoskeleton-wearing walking for new subjects.These results suggest that the proposed approach can improve locomotion mode prediction accuracy for new subjects,thereby ensuring reliable human motion intention prediction in lower-limb exo⁃skeletons.
作者
俞志鹏
王美玲
王成军
凌六一
金力
YU Zhi-peng;WANG Mei-ling;WANG Cheng-jun;LING Liu-yi;JIN Li(School of Artificial Intelligence,Anhui University of Science&Technology,Hefei,Anhui 231131,China;State Key Laboratory of Digital Intelligent Technology for Unmanned Coal Mining,Anhui University of Science&Technology,Huainan,Anhui 232001,China;Anhui Key Laboratory of Bionic Sensing and Advanced Robot technology,Hefei,Anhui 230031,China;Institute of Intelligent Machinery,Hefei Institutes of Physical Science,Chinese Academy of Sciences,Hefei,Anhui 230031,China)
出处
《电子学报》
北大核心
2025年第6期1969-1978,共10页
Acta Electronica Sinica
基金
安徽理工大学高层次引进人才科研启动基金(No.2023yjrc111)
仿生感知与先进机器人技术重点实验室开放基金(No.AHFS2024KF06)。
关键词
下肢外骨骼
下肢运动模式预测
表面肌电信号
迁移学习
多传感器信息融合
lower limb exoskeleton
locomotion mode prediction
surface electromyography
transfer learning
multi-sensor information fusion
作者简介
俞志鹏,男,1995年3月出生于安徽省淮南市.博士,现为安徽理工大学人工智能学院讲师.主要研究方向为人机交互与康复机器人.E-mail:pengyz@aust.edu.cn;王美玲,女,1986年1月出生于安徽省安庆市.博士,现为中国科学院合肥物质科学研究院智能机械研究所副研究员.主要研究方向为机器人与智能装备.E-mail:mlwang@iamt.ac.cn;通讯作者:凌六一,男,1980年7月出生于安徽省枞阳县.博士,现为安徽理工大学人工智能学院教授,博士生导师.主要研究方向为检测技术与智能信息处理.E-mail:lyling@aust.edu.cn。
