Locomotion Intention Prediction via Multi-Sensor Fusion and Transfer Learning for Lower Limb Exoskeletons

YU Zhi-peng; WANG Mei-ling; WANG Cheng-jun; LING Liu-yi; JIN Li

doi:10.12263/DZXB.20241134

您当前的位置：

首页 >

文章列表页 >

Locomotion Intention Prediction via Multi-Sensor Fusion and Transfer Learning for Lower Limb Exoskeletons

PAPERS | 更新时间：2025-10-16

- Locomotion Intention Prediction via Multi-Sensor Fusion and Transfer Learning for Lower Limb Exoskeletons
- ACTA ELECTRONICA SINICA Vol. 53, Issue 6, Pages: 1969-1978(2025)
- 作者机构：
  
  1.安徽理工大学人工智能学院，安徽合肥 231131
  2.安徽理工大学煤炭无人化开采数智技术全国重点实验室，安徽淮南 232001
  3.仿生感知与先进机器人技术安徽省重点实验室，安徽合肥 230031
  4.中国科学院合肥物质科学研究院智能机械研究所，安徽合肥 230031
- 作者简介：
- 基金信息：
  
  Scientific Research Foundation for High-level Talents of Anhui University of Science and Technology(2023yjrc111);Anhui key Laboratory of Bionic Sensing and Advanced Robot Technology(AHFS2024KF06)
- DOI：10.12263/DZXB.20241134
  CLC： TP24;TP391
- Received：18 December 2024，
  
  Revised：2025-05-15，
  
  Published：25 June 2025
- 稿件说明：
移动端阅览
俞志鹏, 王美玲, 王成军, 等. 基于多传感器信息融合和迁移学习的下肢外骨骼运动意图预测研究[J]. 电子学报, 2025, 53(06): 1969-1978.

YU Zhi-peng, WANG Mei-ling, WANG Cheng-jun, et al. Locomotion Intention Prediction via Multi-Sensor Fusion and Transfer Learning for Lower Limb Exoskeletons[J]. Acta Electronica Sinica, 2025, 53(06): 1969-1978.
俞志鹏, 王美玲, 王成军, 等. 基于多传感器信息融合和迁移学习的下肢外骨骼运动意图预测研究[J]. 电子学报, 2025, 53(06): 1969-1978. DOI：10.12263/DZXB.20241134

YU Zhi-peng, WANG Mei-ling, WANG Cheng-jun, et al. Locomotion Intention Prediction via Multi-Sensor Fusion and Transfer Learning for Lower Limb Exoskeletons[J]. Acta Electronica Sinica, 2025, 53(06): 1969-1978. DOI：10.12263/DZXB.20241134

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.

Abstract

Lower limb exoskeletons require the capability to identify the user’s lower-limb motion intentions to provide 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 prediction 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 inter-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 fully 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 exoskeletons.

关键词

Keywords

references

LEE H , FERGUSON P W , ROSEN J . Chapter 11 lower limb exoskeleton systems: Overview [J ] . Wearable Robotics , 2020 , 1 : 207 - 229 .

LI L L , CAO G Z , LIANG H J , et al . Human lower limb motion intention recognition for exoskeletons: A review [J ] . IEEE Sensors Journal , 2023 , 23 ( 24 ): 30007 - 30036 .

LI Z J , LI X , LI Q J , et al . Human-in-the-loop control of soft exosuits using impedance learning on different terrains [J ] . IEEE Transactions on Robotics , 2022 , 38 ( 5 ): 2979 - 2993 .

REN X Q , LI Z J , ZHOU M C , et al . Human intention-aware motion planning and adaptive fuzzy control for a collaborative robot with flexible joints [J ] . IEEE Transactions on Fuzzy Systems , 2023 , 31 ( 7 ): 2375 - 2388 .

UCHITOMI H , HIROBE Y , MIYAKE Y . Three-dimensional continuous gait trajectory estimation using single Shank-Worn inertial measurement units and clinical walk test application [J ] . Scientific Reports , 2022 , 12 ( 1 ): 5368 .

ZHANG S Q , PAN J H , LI L . Non-linear changes of lower extremity kinetics prior to gait transition [J ] . Journal of Biomechanics , 2018 , 77 : 48 - 54 .

汪成亮 , 刘艺锣 . 基于震动信号的异常步态识别 [J ] . 电子学报 , 2023 , 51 ( 8 ): 2088 - 2097 .

WANG C L , LIU Y L . Recognition of abnormal gait based on vibration signal [J ] . Acta Electronica Sinica , 2023 , 51 ( 8 ): 2088 - 2097 . (in Chinese)

KIM J , LEE G , HEIMGARTNER R , et al . Reducing the metabolic rate of walking and running with a versatile, portable exosuit [J ] . Science , 2019 , 365 ( 6454 ): 668 - 672 .

BAUD R , MANZOORI A R , IJSPEERT A , et al . Review of control strategies for lower-limb exoskeletons to assist gait [J ] . Journal of Neuroengineering and Rehabilitation , 2021 , 18 ( 1 ): 119 .

DONALDSON B , BAYNE H , BEZODIS N E . Within-subject repeatability and between-subject variability in posture during calibration of an inertial measurement unit system [J ] . ISBS Proceedings Archive , 2021 , 39 ( 1 ): 224 .

LI K X , ZHANG J H , WANG L F , et al . A review of the key technologies for sEMG-based human-robot interaction systems [J ] . Biomedical Signal Processing and Control , 2020 , 62 : 102074 .

KYEONG S , FENG J R , RYU J K , et al . Surface electromyography characteristics for motion intention recognition and implementation issues in lower-limb exoskeletons [J ] . International Journal of Control, Automation and Systems , 2022 , 20 ( 3 ): 1018 - 1028 .

AFZAL T , IQBAL K , WHITE G , et al . A method for locomotion mode identification using muscle synergies [J ] . IEEE Transactions on Neural Systems and Rehabilitation Engineering , 2017 , 25 ( 6 ): 608 - 617 .

FU P W , ZHONG W J , ZHANG Y Y , et al . Predicting continuous locomotion modes via multidimensional feature learning from sEMG [J ] . IEEE Journal of Biomedical and Health Informatics , 2024 , 28 ( 11 ): 6629 - 6640 .

GAO F , LIU G Y , LIANG F Y , et al . IMU-based locomotion mode identification for transtibial prostheses, orthoses, and exoskeletons [J ] . IEEE Transactions on Neural Systems and Rehabilitation Engineering , 2020 , 28 ( 6 ): 1334 - 1343 .

SU B Y , WANG J , LIU S Q , et al . A CNN-based method for intent recognition using inertial measurement units and intelligent lower limb prosthesis [J ] . IEEE Transactions on Neural Systems and Rehabilitation Engineering , 2019 , 27 ( 5 ): 1032 - 1042 .

席旭刚 , 汤敏彦 , 张自豪 , 等 . 融合表面肌电和加速度信号的下肢运动模式识别研究 [J ] . 电子学报 , 2017 , 45 ( 11 ): 2735 - 2741 .

XI X G , TANG M Y , ZHANG Z H , et al . Lower limb motion recognition based on the fusion of sEMG and acceleration signal [J ] . Acta Electronica Sinica , 2017 , 45 ( 11 ): 2735 - 2741 . (in Chinese)

PENG F , ZHANG C , XU B G , et al . Locomotion prediction for lower limb prostheses in complex environments via sEMG and inertial sensors [J ] . Complexity , 2020 , 2020 ( 1 ): 8810663 .

LIU Z J , LIN W , GENG Y L , et al . Intent pattern recognition of lower-limb motion based on mechanical sensors [J ] . IEEE/CAA Journal of Automatica Sinica , 2017 , 4 ( 4 ): 651 - 660 .

ZHENG E H , WANG Q N . Noncontact capacitive sensing-based locomotion transition recognition for amputees with robotic transtibial prostheses [J ] . IEEE Transactions on Neural Systems and Rehabilitation Engineering , 2017 , 25 ( 2 ): 161 - 170 .

LE D , CHENG S H , GREGG R D , et al . Transfer learning for efficient intent prediction in lower-limb prosthetics: A strategy for limited datasets [J ] . IEEE Robotics and Automation Letters , 2024 , 9 ( 5 ): 4321 - 4328 .

ZHANG K G , WANG J , DE SILVA C W , et al . Unsupervised cross-subject adaptation for predicting human locomotion intent [J ] . IEEE Transactions on Neural Systems and Rehabilitation Engineering , 2020 , 28 ( 3 ): 646 - 657 .

ZHOU X , YU Z , CHEN D , et al . Experimental research on follow-up control of lower limb exoskeleton robot; proceedings of the ISMSEE 2022 [C ] // The 2nd International Symposium on Mechanical Systems and Electronic Engineering . Frankfurt : VDE , 2022 : 1 .

李自由 , 赵新刚 , 张弼 , 等 . 基于表面肌电的意图识别方法在非理想条件下的研究进展 [J ] . 自动化学报 , 2021 , 47 ( 5 ): 955 - 969 .

LI Z Y , ZHAO X G , ZHANG B , et al . Review of sEMG-based motion intent recognition methods in non-ideal conditions [J ] . Acta Automatica Sinica , 2021 , 47 ( 5 ): 955 - 969 . (in Chinese)

NEENU R , TITUS G . Comparative Study of CNN and LSTM-Based Hand Gesture Classification Using EMG Signals [M ] // Artificial Intelligence in Telemedicine . Boca Raton : CRC Press , 2023 : 83 - 96 .

SONG Q Z , MA X J , LIU Y L . Continuous online prediction of lower limb joints angles based on sEMG signals by deep learning approach [J ] . Computers in Biology and Medicine , 2023 , 163 : 107124 .

XIONG D Z , ZHANG D H , ZHAO X G , et al . Deep learning for EMG-based human-machine interaction: A review [J ] . IEEE/CAA Journal of Automatica Sinica , 2021 , 8 ( 3 ): 512 - 533 .

SHERSTINSKY A . Fundamentals of recurrent neural network (RNN) and long short-term memory (LSTM) network [J ] . Physica D: Nonlinear Phenomena , 2020 , 404 : 132306 .

HAMDI M M , AWAD M I , ABDELHAMEED M M , et al . Lower limb motion tracking using IMU sensor network [C ] // 2014 Cairo International Biomedical Engineering Conference (CIBEC) . Piscataway : IEEE , 2014 : 28 - 33 .

SANTURKAR S , TSIPRAS D , ILYAS A , et al . How does batch normalization help optimization? [EB/OL ] . ( 2019-08-15 )[ 2025-05-10 ] . https://arxiv.org/abs/1805.11604v5 https://arxiv.org/abs/1805.11604v5 .

AGARAP A . Deep learning using rectified linear units (relu) [EB/OL ] . ( 2019-02-27 )[ 2025-05-10 ] . https://arxiv.org/abs/1803.08375 https://arxiv.org/abs/1803.08375 .

HAHN S , CHOI H . Understanding dropout as an optimization trick [J ] . Neurocomputing , 2020 , 398 : 64 - 70 .

WEI C F , WANG H , HU F , et al . Single-channel surface Electromyography signal classification with variational mode decomposition and entropy feature for lower limb movements recognition [J ] . Biomedical Signal Processing and Control , 2022 , 74 : 103487 .

GU H , LI W D , LI J . Knee and ankle kinematics in different walking conditions [C ] // 2016 International Conference on Advanced Robotics and Mechatronics (ICARM) . Piscataway : IEEE , 2016 : 266 - 270 .

LI Z J , ZHANG T , HUANG P B , et al . Human-in-the-loop cooperative control of a walking exoskeleton for following time-variable human intention [J ] . IEEE Transactions on Cybernetics , 2024 , 54 ( 4 ): 2142 - 2154 .

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Hybrid Task Assignment for Crowd Sensing Based on Mobility Prediction

Semantic Comparison Model for Binary Programs Based on Statistical Reasoning

Zero Sample Fault Diagnosis Based on Transfer Learning

Radar Target Recognition Method for HRRP Based on Geodesic Flow Kernel

Related Author

YU Zhi-peng

WANG Cheng-jun

LING Liu-yi

JIN Li

JIANG Wei-jin

NIE Cai-yan

LIU Qian

DU Xi-chen

Related Institution

School of Artificial Intelligence, Anhui University of Science & Technology

State Key Laboratory of Digital Intelligent Technology for Unmanned Coal Mining, Anhui University of Science & Technology

Hunan University of Information Technology

College of Frontier Intersection, Hunan University of Technology and Business

Xiangjiang Laboratory

⁰