Research On Mental State Recognition Based On The Fusion Of EEG And ECG Information

The quality of the mental state directly affects the level of performance of the individual,especially for those who are engaged in special operations in key positions,the slightest error will cause serious consequences.In response to this problem,the research on the intelligent identification and evaluation system for mental state was carried out to avoid the serious work errors caused by the poor mental state of the staff in special positions.Thanks to the rapid development of brain-computer interface technology and artificial intelligence algorithms,the state recognition research based on EEG signals has become a hotspot in the field of human-computer interaction in recent years.However,there are still the following problems in the recognition of mental states: First,there is no unified judgment standard for the mental state subjective test,which requires researchers to design themselves,and there are large differences between them;then,the design of the mental state evaluation algorithm has no fixed reference And standards,you can select different types of parameters and combine different algorithms to achieve mental state assessment;finally,the connection between mental state and human physiological parameters is not clearly proposed.There are useless signals in the collected data,and there are too many valid signals The redundant information of the system brings unnecessary consumption to the training of the algorithm model.In response to the above problems,research work has been carried out from the following three aspects:First,the "gold standard" EEG signals for human mental fatigue assessment and ECG signals that accurately reflect the physiological state of the human body were selected as the data basis for this study.Design fatigue and emotion-inducing experiments separately.Relying on the relatively mature judgment criteria of the two to determine the subjective evaluation index of mental state;then,according to the characteristics of the signal,the filter and wavelet transform are used to denoise the signal.The pan-tompkins algorithm and the maximum detection combined detection algorithm are used to locate the ECG signal and extract its time-domain features.The wavelet packet transform is used to decompose the EEG signal into rhythm waves of different frequency bands to extract its energy features.Fatigue identification using BP neural network with additional momentum method overcomes the shortcomings of low learning rate of BP neural network model and easy to fall into local optimal solution.At the same time,non-linear support vector machines are used to classify emotions.Combined with regression analysis theory,a variable parameter mental state evaluation equation is proposed to realize the classification and recognition of mental states;finally,the results of mental state evaluation under key lead and 15 leads are compared and analyzed to find out the brain electrical leads strongly related to mental state.The results of the study show that an algorithm model designed with a combination of fatigue and emotion can effectively evaluate mental states.Has a good performance in multiple experimental tests.At the same time,identifying the leads that are closely related to mental state provides a basis for the study of reconstructing multi-lead EEG signals with a few key leads.