The Research Of Car Following Model Based On Real-time Road Surface Adhesion Condition

Friction coefficient could be used to judge the road surface adhesion condition as important parameter between the vehicle tire and road surface. It is mainly used in the ABS and ESC of vehicles. Along with the development of automation and vehicle sensors, estimation accuracies of friction coefficient and road surface condition are improved. The friction coefficient determines the theoretical maximum deceleration which determines the safe following distance. Different road surface have different friction coefficient. Therefore, considering the influence of real-time road surface condition on car-following has important significance.Firstly, the real-time estimation of friction coefficient is completed by longitudinal vehicle model combining with the data of vehicle sensors and GPS. The slip rate and the normalized traction force are calculated with the data of vehicle sensors and GPS. The slip-slope is fitted using a recursive least-squares parameter identification algorithm and the friction coefficient is estimated through the linear relationship between the slip rate and the normalized traction force in the region of little slip rate.Secondly, the experiment is carried out with differential GPS to investigate and analyze the car-following behavior of urban road and highway. We survey the normal driving on the highway, the normal driving, braking to park and accelerate to launch on the urban road. Then, the car-following difference of a same driver in the different situation is analyzed from the intensity of acceleration and deceleration, the headway and the collision time etc. The car-following difference of different drivers in the same situation is analyzed also.Thirdly, the new car-following model is established by modifying intelligent driver model. The desired car-following gap is modified with the relation between following velocity and headway and the theoretical maximum deceleration which could distinguish the road surface adhesion condition. The construction of the intelligent driver model is optimized according to its disadvantage during high-speed movement.Finally, the new model is simulated in the normal road condition and changing road condition to test the stability and adaptability. The simulation results show that the new model controls vehicle moving securely and steady with the smaller following gap and headway than the intelligent driver model. The new model drive more smoothly than the intelligent driver model with the smaller following gap. It is significant to enhance traffic capacity on basis of guaranteeing the security and comfort of passengers. It is also certified that the new model has better robustness and adaptability for random perturbations than the intelligent driver model.