Continuous and simultaneous estimation of lower limb multi-joint angles from sEMG signals based on stacked convolutional and LSTM models
作者：
Highlights：
• Lower limb joint angle estimation of daily movements is realized using sEMG signals.
• A method is proposed to calculate joint angles using Euler angles measured by IMU.
• Time–frequency analysis of multi-channel sEMG signals is performed in the gait cycles.
• TD features outperform FD or TFD features in lower limb multi-joint angle estimation.
• Conv-LSTM extracts spatiotemporal information and obtain high estimation performance.
摘要
•Lower limb joint angle estimation of daily movements is realized using sEMG signals.•A method is proposed to calculate joint angles using Euler angles measured by IMU.•Time–frequency analysis of multi-channel sEMG signals is performed in the gait cycles.•TD features outperform FD or TFD features in lower limb multi-joint angle estimation.•Conv-LSTM extracts spatiotemporal information and obtain high estimation performance.
论文关键词：Regression analysis,EMG,Joint angle estimation,Deep learning,Human-exoskeleton interaction,Kinematic model
论文评审过程：Received 31 January 2021, Revised 20 January 2022, Accepted 25 April 2022, Available online 5 May 2022, Version of Record 13 May 2022.
论文官网地址：https://doi.org/10.1016/j.eswa.2022.117340
