Motion Estimation for the Control of Upper Limb Wearable Exoskeleton Robot with Electroencephalography Signals

Abstract

Brain-Machine Interface (BMI) has emerged as a powerful tool for assisting disabled people. In this work, we propose a motion estimation method using electroencephalography (EEG) signals to augment human performance. Because the EEG signal occurs before the actual motion is executed, there is a time lag between motion and EEG signals. In this paper, we introduce this time lag to construct a linear model that correlates the electromyography (EMG) signal to the EEG signals based on motion-related features extracted from multi-location EEG signal measurements by Independent Component Analysis (ICA). The constructed model is used to estimate the human muscular activity of shoulder joint from EEG signals. Furthermore, we also discuss the effect on the estimation results with different training data and overlap rates for the model, and finally we know how to select the optimal values of parameters for proposed method. The proposed approach is experimentally verified. Our results suggest that the estimation of EMG signal based on EEG signals is feasible, and demonstrate the potential of using EEG signals via the control of brain-machine interface to support human activities.