Research On Fatigue Driving Detection Method Based On Smartwatch

Fatigue is a physiological phenomenon that occurs frequently and is unavoidable for drivers.Fatigue driving seriously endangers the safety of the lives and property of individuals and others.It has become a very serious social problem,related to personal life and health and property safety,which has attracted great attention from many countries.The existing fatigue driving detection technology has the following problems: the contact signal acquisition method and the complexity of professional measurement equipment make the signal monitoring in the vehicle driving environment inconvenient;the signal monitoring is susceptible to vehicle movement,road conditions,environmental factors,etc.The reliability of the test results is not high;the collected information content relates to the privacy of users.These problems seriously hinder the application of existing fatigue driving detection technology in real life.Mobile wearable devices are expected to provide a more convenient and practical driving fatigue detection method.The perception and calculation of personal mobile devices solve the problems of inconvenience of information monitoring and user privacy.However,due to the simple type of embedded sensors in commercial mobile smart devices,it is not possible to directly obtain physiological characteristic information related to fatigue assessment.Therefore,how to mine potential individual physiological information characteristics through monitoring data is a problem that needs to be solved to ensure the reliability of detection;Secondly,due to the poor environmental adaptability of a single type of signal detection result,and considering the user privacy issue,when combining driver fatigue behavior characteristics for driving fatigue detection,how to detect the physiological characteristics of driver fatigue through non-sensitive data monitored by mobile devices is a problem that needs to be solved to enhance the robustness of the detection results;Finally,due to the limited power of mobile terminals and the high power consumption of sensors used to monitor signals,the device cannot monitor driver's driving fatigue for a long time,so how to scientifically design the startup mechanism of high-power sensors to reduce system power consumption is what we need to solve The problem.This research focuses on the above problems.First,based on the non-sensitive data collected by mobile devices,during the extraction of fatigue features,a heart rate variability feature extraction method based on statistical analysis is designed to obtain RRI(RR interval,the time interval between two adjacent R waves in the electrocardiogram)from the heart rate data,and extract the time domain and frequency domain features of heart rate variability,which solves the problem of difficulty in monitoring heart rate variability features.At the same time,a relative motion detection algorithm for the driver's hand is designed to reduce interference in the extraction of steering wheel motion features;Secondly,taking yawn as a representative of typical individual fatigue physiological behavior,in-depth study and mining the relationship between heart rate data and yawn behavior,a heart rate yawn matching algorithm is proposed to detect the driver's yawning behavior,which improves the accuracy of fatigue detection while enhancing the robustness;Finally,for the problem of power consumption of the system,a universal driving posture detection algorithm is proposed to control the high-power heart rate sensor,which greatly reduces the power consumption of the system and enables the system to meet the driver's long-term fatigue driving detection.After experimental verification,the detection accuracy of this research design method can reach 97.1%,and the energy consumption can be reduced by 33%.The experimental results show that the research can provide new ideas and useful references for fatigue detection research.