A Robot Control Method based on Motor Imagery EEG Signals

In recent years, brain-computer interface (BCI)-based research has become a research direction that has attracted much attention in the field of artificial intelligence, especially the brain-computer interface-based human-computer interaction system has a wide range of development forward. In this paper, a new recognition method is proposed for the problem of low accuracy of multi-classification recognition of EEG signals based on motor imagery, and the method is combined with Nao robot control to realize a human-computer interaction system based on brain-computer interface. In this study, the EEG data collected by EMOTIV EPOC FlEX was used as the experimental basis, and firstly, the wavelet packet decomposition was used to process the EEG signals, extract the four important frequency bands of EEG signals, namely, Alpha, Beta, Theta, and Gamma, and compute the differential entropy (DE) features of each band respectively, and reconstructed the three-dimensional according to the spatial characteristics of the brain electrodes EEG feature cube, and input the features into the convolutional neural network of the visual geometry group network for four classifications. Finally, the robot is controlled to realize different limb movements according to different classification results, thus realizing the function of human-machine interaction system and laying a technical foundation for intelligent control based on EEG signals.