Research On Driver Fatigue Recognition Method Based On Physiological Signals

Fatigue driving the vehicle led to the root cause of traffic accidents, accurate monitoring of fatigue state motorists, and real-time early warning is an important way to prevent traffic accidents, so real-time detection of car driver fatigue has become a concern of experts and scholars at home and abroad. at present, at home and abroad in the exploratory stage of fatigue monitoring method, there is no standardized system and good solutions, fatigue monitoring products already exist mostly poor accuracy and poor comfort, and expensive, Thus, in-depth and meticulous research driver fatigue monitoring technology is very important. Comprehensive application of modern testing, signal analysis, computers, machinery and other technology to develop low-cost, high-performance, high accuracy car driver fatigue prediction and alert system, not only can reduce traffic accidents the benefit of mankind, but also to seize the huge domestic and foreign markets,it has broad prospects for development, huge economic and social benefits.First, I read a lot of literature, analyzed the physiological signal monitoring human fatigue feasibility, and its application in the driver fatigue monitoring. In order to find driver fatigue and driver back pressure distribution, as well as the relationship between heart rate, pulse, fatigue and wakefulness experiments were carried out experiments. To make the experimental data closer to the real situation, experiments in simulated driving platform environment were carried out 16 experiments fatigue and fatigue by the method of subjective evaluation of SD, in experiments carried out while the fatigue state records.Performed using a surface pressure sensors, ECG sensor and pulse sensor signals based on the acquisition method, and were used in time domain analysis, Fourier analysis and a method of chaotic time series of physiological characteristic feature extraction, by the detected data analysis of physiological signals, support vector machine approach to pattern recognition and fatigue level determination, combined with subjective scoring of the test results for statistical analysis, Final results show the effectiveness of the extracted fatigue characteristics and SVM method.