Research And Integrated Implementation Of Hybrid Brain Computer Interface Based On EEG And EMG

As a way of information exchange between brain and external devices,brain computer interface(BCI)plays an important role in numerous fields such as assistant medical assistant.Due to the low recognition rate of EEG signals based on single mode,the hybrid BCI which integrates EEG signals and other physiological signals has become a research focus of BCI.In this thesis,the research of hybrid BCI based on electromyogram(EMG)and electromyogram(EEG)is carried out,which has important theoretical significance and practical application value.Based on the analysis of the research status of hybrid BCI at home and abroad,the design of overall framework of human-computer interaction system of hybrid BCI based on EEG and EMG is completed.EEG and EMG signals are collected by Emotiv sensor.The processing methods of EEG and EMG signals are analyzed,and the deep learning method is selected as the processing methods of EEG and EMG signals.Aiming at the low recognition rate of EEG signals,a method of EEG recognition based on convolutional neural networks(CNN)in time-frequency domain is proposed.This method constructs a two-dimensional graph in time-frequency domain by short-time fourier transform(STFT)of EEG signals,and obtains the input of convolution neural network.A novel structure of CNN is designed by one-dimensional convolution method,and feature extraction is completed.Then classification and recognition are realized by using support vector machine(SVM).The experimental results show that the EEG recognition rate of the proposed method is higher than that of other method such as common space pattern(CSP).In the recognition of EMG signals,deep belief network(DBN)is applied to extract the features of EMG signals,and SVM is used for classification.In the fusion of EEG and EMG,the information fusion method based on SVM and D-S evidence theory is adopted to realize the fusion of EEG and EMG at decision-making level.The experimental results show that the recognition rate of this fusion method is higher than that of other fusion methods.Finally,a human-computer interaction system of hybrid BCI based on EEG and EMG is constructed on the intelligent wheelchair platform,and the hardware and software design of the system is completed.System test shows that the system caneffectively control the movement of intelligent wheelchair.The proposed method is effective,and the human-computer interaction system of hybrid BCI based on EEG and EMG has better stability than that of single mode BCI.