To address the limitation that lower-limb exoskeletons cannot adapt to personalized motion demands, a continuous knee joint angle prediction method based on sEMG (surface electromyography) was proposed and evaluated. The limitation arises from the difficulty of accurately perceiving human movement intention during human-robot interaction rehabilitation tasks. Firstly, the collected sEMG signals were filtered, and features were extracted using a sliding time-window method. Secondly, a prediction model named GAPSO-BP was developed by combining GA (genetic algorithm) and PSO (particle swarm optimization) to optimize BP (back propagation) neural network. The predicted angles were then compared with the actual angles. Finally, a knee exoskeleton human-robot interaction platform was also built using a mirror-control experimental scheme, and system validation was conducted through motion-tracking and assistance-effectiveness tests. Results show that the prediction accuracy of the GAPSO-BP model increases by 4%, 6%, and 7% compared with BP, RBF (radial basis function), and ELM (extreme learning machine) models. In motion-tracking evaluation, the average coefficient of R-Squared, root mean square error, and normalized root mean square error are 91.61%, 2.029°, and 7.328%. In assistance-effectiveness evaluation, the average assistance efficiency is 37.27%. The proposed GAPSO-BP algorithm improves prediction accuracy and provides a reliable human-robot interaction approach for lower-limb exoskeleton systems.
ZHANG Xing, WANG Ning, YANG Ming-xing, LI Jie, YE Ye.
Continuous Prediction of Human Knee Joint Angles Based on Surface Electromyography Signals and Experimental Validation[J].
Science Technology and Engineering, 2026, 26(26): 11367-11376 DOI:10.12404/j.issn.1671-1815.2507077
选取4名健康被试参与人机交互在线实验,在左腿的股直肌、股二头肌和股外侧肌处布置sEMG传感器作为数据采集点,右腿穿戴膝关节助力外骨骼进行人体助力辅助,4名被试的人机交互实验如图11所示。实验分别在1.0、1.3、1.5 km/h 3种速度下进行,全程采集被试穿戴外骨骼进行助力运动过程中的sEMG、关节角度和跟踪误差等运动数据。
3.3.2 运动跟随效果
采用决定系数(R2)、均方根误差(root mean square error,RMSE)及标准化均方根误差(normali-zed root mean square error,NRMSE)作为在线人机交互实验中运动跟随效果的定量评估统计指标。当R2越接近1且RMSE、NRMSE越小时,表明运动跟随效果越好,相关表达式为
MaX, LüL, LiuW, et al. Exoskeletons in the context of soldiers: current status and future research trends[J]. Biomimetic Intelligence and Robotics, 2025, 5(3): 100254.
LiuRuiping, OuyangJun. Development andapplication of lower limb exoskeleton rehabilitation robot[J]. Journal of Sun Yat-sen University(Medical Sciences Edition), 2023, 44(2): 354-360.
[4]
Sczesny-KaiserM, TrostR, AachM, et al. A randomized and controlled crossover study investigating the improvement of walking and posture functions in chronic stroke patients using HAL exoskeleton: the HALESTRO study (HAL-Exoskeleton STROke Study)[J]. Frontiers in Neuroscience, 2019, 13: 259.
ZhangMinglu, ZhongDaofang, TianYing, et al. Research status of wearable lower extremity exoskeleton robots[J]. Science Technology and Engineering, 2021, 21(19): 7856-7862.
[7]
SongT, YanZ, GuoS, et al. Review of sEMG for robot control: techniques and applications[J]. Applied Sciences, 2023, 13(17): 9546.
HuangHaifeng, LiuPeisen, LiQing, et al. Review: intelligent control and human-robot interaction for collaborative robots[J]. Chinese Journal of Engineering, 2022, 44(4): 780-791.
[10]
HandelmanD A, OsbornL E, ThomasT M, et al. Shared control of bimanual robotic limbs with a brain-machine interface for self-feeding[J]. Frontiers in Neurorobotics, 2022, 16: 918001.
[11]
LiZ, YuanY, LuoL, et al. Hybrid brain/muscle signals powered wearable walking exoskeleton enhancing motor ability in climbing stairs activity[J]. IEEE Transactions on Medical Robotics and Bionics, 2019, 1(4): 218-227.
YouBo, LiuJiaqi, ChengChenchen, et al. Human-computer interaction follow-up control based on sEMG[J]. Chinese Journal of Scientific Instrument, 2025, 46(3): 123-142.
SongYongxian, WangXiangxiang, LiYuanyuan, et al. Lowerlimb joint moment prediction method based on kernel limit learning machine[J]. Science Technology and Engineering, 2024, 24(11): 4599-4606.
[16]
DingQ C, XiongA B, ZhaoX G, et al. A review on researches and applications of sEMG-based motion intent recognition methods[J]. Acta Automatica Sinica, 2016, 42(1): 13-25.
[17]
ForoutanniaA, AkbarzadehA. A deep learning strategy for EMG-based joint position prediction in hip exoske-leton assistive robots[J]. Biomedical Signal Processing and Control, 2022, 75: 103557.
LiWenju. Research on human-computer interaction strategies and methods for lower limb exoskeleton based on EEG and sEMG[D]. Daqing: Northeast Petroleum University, 2024.
[20]
LuC, QiQ, LiuY, et al. Exoskeleton recognition of human movement intent based on surface electromyographic signals[J]. IEEE Access, 2024, 12: 53986-54004.
[21]
LiZ Y, ZhaoX G, ZhangB, et al. Review of sEMG-based motion intent recognition methods in non-ideal conditions[J]. Acta Automatica Sinica, 2021, 47(5): 955-969.
[22]
XiongD, ZhangD, ZhaoX, et al. Deep learning for EMG-based human-machine interaction: a review[J]. IEEE/CAA Journal of Automatica Sinica, 2021, 8(3): 512-533.
[23]
WuQ, ChenY. Adaptive cooperative control of a soft elbow rehabilitation exoskeleton based on improved joint torque estimation[J]. Mechanical Systems and Signal Processing, 2023, 184: 109748.
[24]
ZhangQ, LambethK, SunZ, et al. Evaluation of a fused sono-myography and electromyography-based control on a cable-drive Anankle exoskeleton[J]. IEEE Transactions on Robotics, 2023, 39(3): 2183-2202.
[25]
ZhuY, WuQ, ChenB, et al. Physical human-robot interaction control of variable stiffness exoskeleton with sEMG-based torque estimation[J]. IEEE Transactions on Industrial Informatics, 2023, 19(10): 10601-10612.
LiuKeping, SunRuiling, ChaiYuanyuan, et al. Upper limb joint angles estimation based on general regression neural network[J]. Science Technology and Engineering, 2021, 21(17): 7187-7192.
TangLu, YangXilin, WangXiangrui, et al. An attention mechanism-based CNN-LSTM framework for lower limb knee joint angle prediction[J]. Journal of Data Acquisition and Processing, 2024, 39(4): 996-1008.
WangHai, TaoQing, SuNa, et al. Method for accurate sensing of finger joint angles using surface electromyography signals[J]. Journal of Xi’an Jiaotong University, 2022, 56(8): 160-167.
[32]
WangC, HeB, WeiW, et al. Prediction of contralateral lower-limb joint angles using vibroar thrography and surface electromyography signals in time-series network[J]. IEEE Transactions on Automation Science and Engineering, 2022, 20(2): 901-908.
[33]
ZhengK, LiuS, YangJ, et al. SEMG-based continuous hand action prediction by using key state transition and model pruning[J]. Sensors, 2022, 22(24): 9949.
WangRenzhi, ZhangWeiguo, KouMiaomiao, et al. Prediction of deep excavation settlement based on improved NNA and BP neural network model[J]. Science Technology and Engineering, 2025, 25(24): 10416-10425.
[36]
WeiQ, LiZ, ZhaoK, et al. Synergy-based control of assistive lower-limb exoskeletons by skill transfer[J]. IEEE/ASME Transactions on Mechatronics, 2019, 25(2): 705-715.