Research On Psychological State Judgment Model Based On Brain Computer Interface

Human mental activity is a manifestation of advanced functions,including cognition,mental state and volitional behavior.The psychological state assessment is based on the participants' psychological changes or other reactions to analyze their psychological state at that time.With the development of brain science,signal processing technology and scientific calculation methods,psychological testing technology has developed into a comprehensive discipline including physiology,psychology,electronics and computer technology.At present,the technology has been widely used in psychological testing of mental illness,health monitoring,diagnosis and treatment.In this paper,music videos with different characteristics are used as stimulating materials to induce different psychological states of subjects and collect their brain electrical signals(EEG).By extracting different features of EEG and using different classification recognition algorithms,the brain electrical signals and humans The relationship between different psychological states was studied and a set of mental state judgment models was established.First,32-channel EEG acquisition equipment is used to collect human brain signals,and the collected signals are filtered through a band-pass filter.Secondly,the power spectral density and wavelet energy entropy are introduced separately,and these two algorithms are used to extract features of EEG signals.Then,three methods are used to classify the extracted signals: ELM extreme learning machine,CNN convolutional neural network and ANN artificial neural network;finally,a mental state judgment model is created and verified.Through the design experiments,it is proved that the recognition rate of the two extraction algorithms is low in the feature extraction stage.We combine these two algorithms into a new feature vector to extract features and find that the recognition rate is greatly improved.Compared with the three classification recognition algorithms,it is found that in most cases,the classification effect of ELM algorithm is generally higher than the other two,and the recognition rate is 78.23% when the two feature extraction algorithms are combined.After studying the test results,we can find that the mental state judgment model has a higher classification and recognition rate and is helpful for psychological research.