Citation :

Tri Nguyen, Dinh-Hung Tran, Thi-Duyen Bui, Tran-Thang Le, Ngoc-Khoat Nguyen, "Novel Surface Electromyography-based Movement Intent Identification Method using Kalman Filtering and K-Nearest Neighbors to Control an Exoskeleton Robot," International Journal of Electrical and Electronics Engineering, vol. 13, no. 8, pp. 1-10, 2026. Crossref, https://doi.org/10.14445/23488379/IJEEE-V13I8P101

Abstract

The purpose of this paper is to identify a novel hybrid control architecture for exoskeleton robot movement using Surface Electromyography (sEMG) signals. It addresses instability and ‘label flickering’ in machine learning by using a parallel set of 54 independent Kalman-filtered features applied directly in the feature space. Based on 9-channel data from the SIAT-LLMD dataset, this work extracts 6 time-domain features using 200 ms sliding windows, including: RMS, MAV, WL, SSI, DASDV and ZC. After extracting meaningful features, the signals are smoothed using an efficient 1D Kalman filter. The result is a 64.7% reduction of noise for training the K-Nearest Neighbors (KNN) model with the factor k equals 1. From significant simulation results, the proposed engineering method meets the real-time requirements of a lower extremity exoskeleton robotic control system by achieving 99.1% classification accuracy on a 16-class test dataset with execution times under 100ms per window. To assess the control relevance of the decoded intent, an sEMG-modulated Fuzzy-PI controller was compared with PID and boundary-layer Sliding-Mode Control (SMC) on a three-degree-of-freedom lower-limb exoskeleton under parameter uncertainty, actuator constraints, and load disturbances. Fuzzy-PI attained a mean tracking RMSE of 2.977°, corresponding to reductions of 48.6% and 46.1% relative to PID and SMC, respectively, without a material increase in RMS actuator torque. The combined recognition-and-control framework, therefore, provides a computationally efficient route from muscle activity to stable assistive motion. The proposed algorithm offers a good balance for embedded systems in lower extremity rehabilitation robotic applications, but it is also able to be applied for the upper ones, achieving a good balance between accuracy, stability, and computational cost.

Keywords

Kalman filter, K-Nearest Neighbors, Exoskeleton robot, Movement Intent Recognition, Surface Electromyography.

References

1. Hao Xu, and Anbin Xiong, “Advances and Disturbances in sEMG-based Intentions and Movements Recognition: A Review,” IEEE Sensors Journal, vol. 21, no. 12, pp. 13019-13028, 2021.
   [CrossRef] [Google Scholar] [Publisher Link]  
2. Wei Li, Ping Shi, and Hongliu Yu, “Gesture Recognition using Surface Electromyography and Deep Learning for Prostheses Hand: State-of-the-Art, Challenges, and Future,” Frontiers in Neuroscience, vol. 15, pp. 1-20, 2021.
   [CrossRef] [Google Scholar] [Publisher Link]   
3. Harshavardhana T. Gowda et al., “A Database of Upper Limb Surface Electromyogram Signals from Demographically Diverse Individuals,” Scientific Data, vol. 12, no. 1, pp. 1-10, 2025.
   [CrossRef] [Google Scholar] [Publisher Link]   
4. Athar Ali et al., “Systematic Review of Back-Support Exoskeletons and Soft Robotic Suits,” Frontiers in Bioengineering and Biotechnology, vol. 9, pp. 1-15, 2021.
   [CrossRef] [Google Scholar] [Publisher Link]   
5. Wenhao Wei et al., “Surface Electromyogram, Kinematic, and Kinetic Dataset of Lower Limb Walking for Movement Intent Recognition,” Scientific Data, vol. 10, no. 1, pp. 1-16, 2023.
   [CrossRef] [Google Scholar] [Publisher Link]   
6. Luís Moreira et al., “Lower Limb Kinematic, Kinetic, and EMG Data from Young Healthy Humans During Walking at Controlled Speeds,” Scientific Data, vol. 8, no. 1, pp. 1-11, 2021.
   [CrossRef] [Google Scholar] [Publisher Link]   
7. Emma Reznick et al., “Lower-Limb Kinematics and Kinetics during Continuously Varying Human Locomotion,” Scientific Data, vol. 8, no. 1, pp. 1-12, 2021.
   [CrossRef] [Google Scholar] [Publisher Link]    
8. Jonathan Camargo et al., “A Comprehensive, Open-Source Dataset of Lower Limb Biomechanics in Multiple Conditions of Stairs, Ramps, and Level-Ground Ambulation and Transitions,” Journal of Biomechanics, vol. 119, 2021.
   [CrossRef] [Google Scholar] [Publisher Link]   
9. Yu Luan et al., “HAR-sEMG: A Dataset for Human Activity Recognition on Lower-Limb sEMG,” Knowledge and Information Systems, vol. 63, no. 10, pp. 2791-2814, 2021.
   [CrossRef] [Google Scholar] [Publisher Link]   
10. Hristo Dimitrov, Anthony M.J. Bull, and Dario Farina, “High-Density EMG, IMU, Kinetic, and Kinematic Open-Source Data for Comprehensive Locomotion Activities,” Scientific Data, vol. 10, no. 1, pp. 1-10, 2023.
    [CrossRef] [Google Scholar] [Publisher Link]  
11. Robert V. Schulte et al., “Database of Lower Limb Kinematics and Electromyography during Gait-Related Activities in Able-Bodied Subjects,” Scientific Data, vol. 10, no. 1, pp. 1-10, 2023.
    [CrossRef] [Google Scholar] [Publisher Link]   
12. Junyo Boo et al., “Comprehensive Human Locomotion and Electromyography Dataset: Gait120,” Scientific Data, vol. 12, no. 1, pp. 1-10, 2025.
    [CrossRef] [Google Scholar] [Publisher Link]   
13. Jiashuai Wang et al., “Action Recognition of Lower Limbs based on Surface Electromyography Weighted Feature Method,” Sensors, vol. 21, no. 18, pp. 1-17, 2021.
    [CrossRef] [Google Scholar] [Publisher Link]   
14. Junyao Wang, Yuehong Dai, and Xiaxi Si, “Analysis and Recognition of Human Lower Limb Motions based on Electromyography (EMG) Signals,” Electronics, vol. 10, no. 20, pp. 1-21, 2021.
    [CrossRef] [Google Scholar] [Publisher Link]   
15. Congcong Zhou et al., “Ankle Foot Motion Recognition based on Wireless Wearable sEMG and Acceleration Sensors for Smart AFO,” Sensors and Actuators A: Physical, vol. 331, 2021.
    [CrossRef] [Google Scholar] [Publisher Link]  
16. Xingjian Wang et al., “sEMG-based Consecutive Estimation of Human Lower Limb Movement by using Multi-Branch Neural Network,” Biomedical Signal Processing and Control, vol. 68, 2021.
    [CrossRef] [Google Scholar] [Publisher Link]  
17. Bruno Dutra, Antonio Silveira, and Antonio Pereira, “Grasping Force Estimation using State-Space Model and Kalman Filter,” Biomedical Signal Processing and Control, vol. 70, 2021.
    [CrossRef] [Google Scholar] [Publisher Link]   
18. Kaitlin G. Rabe, Tommaso Lenzi, and Nicholas P. Fey, “Performance of Sonomyographic and Electromyographic Sensing for Continuous Estimation of Joint Torque during Ambulation on Multiple Terrains,” IEEE Transactions on Neural Systems and Rehabilitation Engineering, vol. 29, pp. 2635-2644, 2021.
    [CrossRef] [Google Scholar] [Publisher Link]  
19. Christian Morbidoni et al., “Machine-Learning-based Prediction of Gait Events from EMG in Cerebral Palsy Children,” IEEE Transactions on Neural Systems and Rehabilitation Engineering, vol. 29, pp. 819-830, 2021.
    [CrossRef] [Google Scholar] [Publisher Link]
20. Jingxuan Liu, Qing Tao, and Bin Wu, “Dynamic Muscle Fatigue State Recognition based on Deep Learning Fusion Model,” IEEE Access, vol. 11, pp. 95079-95091, 2023.
    [CrossRef] [Google Scholar] [Publisher Link]  
21. Stefan Haufe et al., “Gait Event Prediction using Surface Electromyography in Parkinsonian Patients,” Bioengineering, vol. 10, no. 2, pp. 1-16, 2023.
    [CrossRef] [Google Scholar] [Publisher Link]
22. Andrea Tigrini et al., “Phasor-Based Myoelectric Synergy Features: A Fast Hand-Crafted Feature Extraction Scheme for Boosting Performance in Gait Phase Recognition,” Sensors, vol. 24, no. 17, pp. 1-15, 2024.
    [CrossRef] [Google Scholar] [Publisher Link]
23. Alessandro Mengarelli et al., “Myoelectric-based Estimation of Vertical Ground Reaction Force during Unconstrained Walking by a Stacked One-Dimensional Convolutional Long Short-Term Memory Model,” Sensors, vol. 24, no. 23, pp. 1-19, 2024.
    [CrossRef] [Google Scholar] [Publisher Link]
24. Ye Ye et al., “Online Pattern Recognition of Lower Limb Movements based on sEMG Signals and its Application in Real-Time Rehabilitation Training,” Robotica, vol. 42, no. 2, pp. 389-414, 2024.
    [CrossRef] [Google Scholar] [Publisher Link]
25. Shing-Hong Liu et al., “Estimating Gait Parameters from sEMG Signals using Machine Learning Techniques under Different Power Capacity of Muscle,” Scientific Reports, vol. 15, no. 1, pp. 1-15, 2025.
    [CrossRef] [Google Scholar] [Publisher Link]