Research On Driver Steering Model Based On Two-Point Preview

With the continuous improvement of people’s living standards, the recoverable amount of the automobile is increasing. However, the traffic safety problem has also been catching people’s eyes.People’s requirements on automobile safety are getting higher and higher. Therefore, how to evaluate vehicle safety performance objectively and fairly has becoming the series problem for the researchers.Due to many real vehicle tests being carried out in extreme conditions, the driver’s life safety issues pose a serious threat. Considering the above factors, this paper, study on the driver’s steering behavior and expect to establish a driver’s steering model in accord with the real driver steering behavior, and then set up an objective and fair evaluation system for the safety performance of automotive. In addition, the research of driver steering model also has a certain reference value for the development of intelligent vehicle steering system.This paper study on the driver steering model and be combined with the vehicle and road model to establish the human-vehicle-road closed-loop control system model. The main content is as follows:First, to establish vehicle-road model. The first step is analyzing vehicle lateral motion stress to establish a linear two degrees of freedom vehicle dynamics model; The second step is creating fit curves used the discrete road data points to build S-shaped curve road model; The third step is establishing vehicle-road relative position model based on vehicle model and road model, that is called vehicle-road model.Then, to establish driver subsystem models. The first step is analyzing the driver’s steering manipulation to find out the involved subsystem models when the vehicle is steering; The second step is analyzing and building mathematical model for the driver subsystem models referring to human science research results, eventually to establish a visual subsystem model based on two-point preview vision mechanism, a yaw acceleration perception model on vestibular system, a steering decision model based on brain emotional learning(BEL) and an arm muscle model.Finally, to establish human-vehicle-road closed-loop control system model and carry on the related simulation analysis. The key to solve the problem is how to combine the vehicle-road and driver model to establish human-vehicle-road closed-loop control system model. The first step is building the single-point preview driver steering model based on the BEL model that is compared to the driver steering model based on the single-point preview optimal curvature. The comparison result verifies the validity of the proposed model and the consistency of the real driver steering behavior; thesecond step is establishing the two-point preview driver steering model contrast with the single-point preview driver steering model based on the BEL model. The visual mechanism of two-point preview is proved better than single-point preview; The third step is to comparing the two-point preview driver steering model based on the BEL model with the multi-point preview driver steering model based on LQR, and the comparison result shows that the former model is inferior to the later. However, the two-point preview driver steering model also can complete the accurate tracking of the target trajectory and has superior properties, which means the gap between the two models is not significant.In conclusion, due to reproduce a real driver steering behavior, the driver steering model’s physical meaning is clear and also can reproduce the real driver of the steering behavior well,meantime, in accord with the reality of the driver’s steering behavior. In addition, because of this model can realize accurate tracking of the target trajectory, there is also a certain reference value for the development of intelligent vehicle steering system.