The Sdudy On The Discriminating Method Of Driving Fatigue Based On Physiological Signal

The traffic accidents occurs frequently because of the increasing cars, whichcaused by driving fatigue occupy a big proportion of the traffic accidents. So, it hasimportant significance that timely judged driving fatigue and effectively preventedand stopped the occurrence of driving fatigue so as to reduce the traffic accidents andimprove road safety.Firstly, the paper focuses on the formation of the driving fatigue, influencefactors and the relationship of driver’s physiological signal and driving fatigue basedon the present research at home and abroad. According to the present defects situationof driving fatigue and the characteristics of physiological signal, it is proposed theidentification method of driving fatigue based on the driver’s physiological signal.Secondly, the experiments were conducted in the simulator because of the safetyand limitations. However, the effectiveness of the simulation modules is worthstudying and discussing based on physical signal, so this paper firstly study thevalidation of the simulation modules. Design the experiment scheme, select the realscene roads, which mainly includes the common roads and special sections (a sharpturn, a steep slope), and build the same scene in the simulation modules. Request thedrivers drive the cars at the same speed and the time in the real world and simulator.Collect the driver’s physiological signal and select the appropriate indexes, thenconduct the comparative analysis and statistics analysis using statistical methods.Thus the relative effectiveness and absolute validity of driving simulation moduleswere obtained, and it laid a foundation for the fatigue experiments based onphysiological signal in the simulation modules.Design the reasonable driving fatigue experiments. The experimental scene isrelatively monotony, so it is effective for the experiments based on physiologicalsignal in driving simulation modules. Drivers were required to drive cars until fatiguein the monotony simulation modules. Collected the driver’s physiological signal,recorded the experiment video and selected driving fatigue time according to thevideo. The effective information were extracted, and proposed the discriminatingmethod of driving fatigue based on EEG entropy and ECG LF clustering value. Thenthe thresholds were obtained using ROC curves. And the additional experiments wereconducted to prove the threshold value of EEG entropy and ECG LF clustering value had high accuracy. On this basis, the driving fatigue judging system was build and thehierarchical discrimination model was formed based on EEG entropy and ECG LFclustering value. It laid a solid theoretical foundation for discrimination practice of thedriving fatigue.The experimental validation was formed for simulator effectiveness, and itprovides the new ideas and methods for the related study. The threshold value ofdriving fatigue is playing a critical role to study driving fatigue and prevent thegeneration of driving fatigue. And it laid a certain research foundation for relatedresearch of dangerous driving behavior.