Application Research Of Functional Brain Network Based On Shortest Path Tree In Fatigue Driving Status Recognition

Aiming at the problem that the accuracy of fatigue driving state recognition model based on EEG is not ideal and the current lack of research on the neural mechanism of fatigue driving,this project is dedicated to the research of constructing an effective functional brain network(FBN)based on the shortest path tree(SPT)for the state recognition of fatigue driving,the fatigue driving state recognition model and the neural mechanism model of fatigue generation of FBN based on SPT are established,the fatigue driving state recognition method and the neural mechanism model of fatigue generation method of FBN based on the SPT were proposed and designed.The main research content and stage research results are as follows:(1)The key technologies of functional brain networks(FBN)were comprehensively analyzed,including the FBN based on SPT(the network is denoted as SPT-FBN).To solve the problem that the density of FBN based on the union of shortest path tree is too high,an improved method is proposed,which is the FBN based on the combined shortest path tree(the network is denoted as CSPT-FBN),the corresponding construction algorithm is designed.(2)The network properties of SPT-FBN and CSPT-FBN are analyzed.The fatigue driving state recognition model based on SPT-FBN and CSPT-FBN and the neural mechanism analysis model based on CSPT-FBN were established,the corresponding algorithm was designed.(3)On the basis of real data,the model and algorithm are tested and analyzed comprehensively.Through test and result analysis,it is concluded that :(a)the fatigue driving identification model of CSPT-FBN constructed based on seven electrodes P4,CPZ,CP4,Cz,FC4,CP3 and C4 can effectively identify fatigue driving;(b)in terms of fatigue driving status recognition,SPT/CSPT-FBN method has higher recognition accuracy than FBN method based on minimum spanning tree or sparsity.(c)under beta wave,leaf fraction,tree hierarchy and kappa change value of SPT-FBN are negatively correlated with the recognition rate of fatigue driving state;(d)there is an abnormality discharge in the beta band of the EEG between the central zone and the central top zone of the right brain,and the occurrence of fatigue is likely to be closely related to the region.Finally,a stage summary of the research work is gave,the shortcomings of the current work and the prospect of the next step are put forward in this paper.