Study On Driving Fatigue Based On EEG Signal

Driving fatigue is one of the main causes of traffic accidents.How to detect whether drivers are driving fatigue is the most important thing to prevent driving fatigue.At present,researchers have proposed many physiological indicators to assess the fatigue level of drivers,and EEG signals are considered reasonable and effective after testing.This thesis mainly studies the detection of driver’s driving state based on eeg signals,and the specific contents are as follows:(1)The existing methods of fatigue detection and fatigue relief are analyzed,and the causes and manifestations of fatigue,the characteristics and acquisition methods of EEG signal,as well as the fatigue analysis technology based on EEG signal are described.(2)The experiment was designed to collect the EEG signals of 10 subjects in the awake state,the fatigue state and the state after releasing the irritating odor.Based on the processing methods and steps of EEG signals,the eeg signals collected above were preprocessed,and features were extracted by time domain,frequency domain and nonlinear dynamics methods.(3)A new improved salp swarm algorithm(E-SSA)was proposed by introducing scale-free network and inertial weight into salp swarm algorithm and enhancing its ability to jump out of local optimal solution.BP neural network is used to select the classifier of pattern recognition.Aiming at the parameter problem of BP neural network,parameters are optimized by combining the feature extraction data and E-SSA,and the best parameter training classifier is obtained.The BCI2003 motor imagery EEG data set is used to test the method,and the results of an example show that the method is feasible and effective.(4)For the redundant channel problem,particle swarm optimization algorithm and ion motion algorithm are used for channel selection,and the optimal channel set and common channel set are obtained.The classification calculation results show that the particle swarm optimization algorithm is better than the ion motion algorithm.Based on the sample entropy topographic map,the characteristic rules of the common channels in three states of wakefulness,fatigue,and release of irritating odors were analyzed,and the thresholds for judging driver fatigue and regaining sobriety were proposed.The results show that the occipital and parietal areas are associated with driving fatigue,and irritating smells(toilet water and balm essence)can alleviate fatigue,which provides a basis for future human-computer interaction and brain-computer interface real-time detection of driver fatigue and selection of measures to alleviate fatigue.