Feature Extraction of Motor Imagination EEG Signals in AR Model Based on VMD
Wulin Zhang, Zeyu Liang, Zirui Liu, Jie Gao · 2021 International Conference on Electrical, Computer and Energy Technologies (ICECET) · 2021
Because the signal-to-noise ratio of Electroen-cephalograph (EEG) signals of motor imagination is low, unstable and significant different, it has a negative impact on EEG recognition. An effective feature extraction algorithm can improve the recognition rate of EEG signals in brain computer system. In this paper, an AR model feature extraction algorithm based on Variational modal decomposition (VMD) is proposed. Firstly, the EEG signal is decomposed into multiple eigenmodal components by VMD, and the required components are selected and estimated by AR model. Finally, the extraction success rate is improved to 84.05%.