Prediction of trunk flexion angle using a TCN-LSTM model based on a temporal pattern attention mechanism

Abstract Lumbar exoskeletons have broad application prospects in intelligent rehabilitation assistance and human-machine interaction. However, their practical deployment is challenged by the complex neuromuscular control mechanisms of trunk muscles and the difficulty of recognizing movement intentions during balance maintenance and postural regulation. To address this problem, this study designed a simulated bending experiment with different external loads and left-hand support conditions as independent variables to acquire surface electromyography (sEMG) signals from lower-back muscles in complex dynamic scenarios. Based on the experimental data, a hybrid TCN-LSTM model with a Temporal Pattern Attention (TPA) mechanism, namely TLT, is proposed to predict trunk flexion angle by modeling nonlinear temporal dependencies and multivariable spatial coupling characteristics in the signals. Experimental results show that, compared with traditional LSTM and CNN-LSTM models, the proposed method improves both the root mean square error (RMSE) and the coefficient of determination (R 2 ), with the average R 2 reaching 0.898 ± 0.024. These findings verify the effectiveness and practical value of the proposed model and provide a reference for precise exoskeleton control.

Authors

Institutions

Publication Details

Journal
Measurement Science and Technology
Published
2026-11-16
DOI
https://doi.org/10.1088/1361-6501/aec23f
Primary Topic
Muscle activation and electromyography studies
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Prediction of trunk flexion angle using a TCN-LSTM model based on a temporal pattern attention mechanism

Zhigang Zou, Qi Wang
1 links
Measurement Science and Technology
Muscle activation and electromyography studies
article

Prediction of trunk flexion angle using a TCN-LSTM model based on a temporal pattern attention mechanism

Zhigang Zou, Qi Wang
article en

Abstract

Abstract Lumbar exoskeletons have broad application prospects in intelligent rehabilitation assistance and human-machine interaction. However, their practical deployment is challenged by the complex neuromuscular control mechanisms of trunk muscles and the difficulty of recognizing movement intentions during balance maintenance and postural regulation. To address this problem, this study designed a simulated bending experiment with different external loads and left-hand support conditions as independent variables to acquire surface electromyography (sEMG) signals from lower-back muscles in complex dynamic scenarios. Based on the experimental data, a hybrid TCN-LSTM model with a Temporal Pattern Attention (TPA) mechanism, namely TLT, is proposed to predict trunk flexion angle by modeling nonlinear temporal dependencies and multivariable spatial coupling characteristics in the signals. Experimental results show that, compared with traditional LSTM and CNN-LSTM models, the proposed method improves both the root mean square error (RMSE) and the coefficient of determination (R 2 ), with the average R 2 reaching 0.898 ± 0.024. These findings verify the effectiveness and practical value of the proposed model and provide a reference for precise exoskeleton control.

Measurement Science and Technology
Shanghai Dianji University (CN)
Good health and well-being
Muscle activation and electromyography studies
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon