Study On Driving Fatigue Based On ECG Signal

With the continuous development of the global economy,the national car ownership continued to increase,the incidence of road traffic accidents have been high.Research shows fatigue driving is one of the important causes of traffic accidents,which seriously threatens the life and property safety of the drivers.Consequently,the fatigue driving real-time detection and early warning are meaningful for avoiding the occurrence of traffic accidents and protecting the drivers' life,furthermore,it can improve the working efficiency for drivers.In this thesis,according to the traditional method of fatigue driving judgment based on ECG signal,the acquisition time is too long and the response speed can not meet the requirements of practical application,and puts forward the method for analyzing driver fatigue state based on short-time ECG signal,solves the problem of extracting the HRV(Heart Rate Variability)frequency domain indexes in the short-time ECG signal,establishes the fatigue identification model based on the support vector machine(SVM),to achieve the use of short-time ECG signal quickly and accurately determine the driver's fatigue state of the target.The main research contents of this thesis are as follows:Firstly,in order to collect the short-time ECG signal,the ECG acquisition device is designed to collect ECG signal every 5 seconds,and collected the driver's ECG data samples by driving simulation experiment.Then,the ECG data samples were pretreated by wavelet denoising,and the adaptive differential threshold method was used to detect the R wave,and the frequency domain analysis was performed by CZT(Linear frequency modulation Z transform)algorithm.HRV time-frequency domain characteristic index:(1)Time domain index: SDNN(standard deviation of RR interval);(2)Frequency domain index: LFnorm(normalized low-frequency power),HFnorm(normalized high-frequency power),LF/HF(The ratio of low-frequency power to high-frequency power).Finally,the fatigue driving recognition model based on support vector machine was constructed.The ECG data samples were used as the training set and test set in the recognition model,each sample vector contains SDNN,LFnorm,HFnorm,LF/HF these 4 characteristic parameters.The SVM(Support Vector Machines)optimal parameters c=4 and g=16 were selected by cross validation.In the experiments conducted,the recognition results show that the normal and tired are 86.5% and 82.5% respectively.According to the good performance of the classifier and the higher classification accuracy,it can be seen that the HRV time-frequency domain characteristic index of fatigue driving extracted from the short-time ECG signal can effectively identify the fatigue driving state.The research of this thesis solves the real-time problem of fatigue driving detection method based on ECG signal from the angle of shortening the length of ECG signal acquisition,provides effective theoretical support for real-time detection of driver fatigue state,and it is also helpful to promote the development of fatigue driving detection to real vehicle application.