The Research About Drive’s Skill Level Based On Eye Movemen And Operating Signal

With the increase of private car,the incidence of traffic accidents is becoming more and more high,which the ratio that caused by the driver’s improper operation could reach 80%.Therefore,in order to help drivers complete driving task better,more and more driving assistance system have been developed,which can reduce the number of traffic accidents.However,when intelligent vehicle driving assistance system was developed,developers make the situational awareness on the surrounding objects without considering the driver himself.When driving assistance system under the same environment,the control strategy is exactly the same,not changing the control strategy depending on the personnel driving proficiency.Based on this,this paper proposes using the driver’s operation signals as well as eye movement signal to evaluate driver’s skills,and to put forward the driver’s skill evaluation model,this paper mainly completed the following works:(1)Set up the Driver-in-Loop simulation experiments and design、built the experimental scenes.In driver-in-Loop simulation platform based on Carsim and Labview,building the virtual scene and when drivers drive,they need to put on the eye movement device where the eye movement information and the operating signals(steering wheel Angle,lateral acceleration,and speed)are collected.The real-time data gathered by Data Acquisition Card is transferred from the Real-Time terminal to the host by data sharing and save in the host.Therefore,a database can be established by these data,and the driving skill models can be trained offline by the database.(2)Establish driving skill evaluation model based on driver operation signal and eye movement signal.Many drivers’ skill recognition methods have been summarized and the advantages and disadvantages of SVM have been analyzed.First,in order to reduce the error rate,using the driver ’s eye movement signal to label driver ’s skill.Then,using the driver ’s operation signal(Speed,Steering wheel Angle and Lateral acceleration)to identify the driver’s skill.And compare the signals using a single operating and multiple signals which combined different operation signals.The results showed that using a combination of speed and lateral acceleration as characteristic,the highest accuracy could be 90%.(3)Driving skill evaluation model is set up based on operation signal.Using standard function to build driving skill evaluation model,it is concluded that using exponential function and liner function could have better performance.Thus,using them to combine,the result showed that using(0.3 × exp(6 × V)+ 0.7 ×(8 × Ac)to evaluate driver’s skill could have better performance with speed and lateral acceleration.These also verify the result of the second part.