In order to explore the adaptability of the existing merging area in highway to autonomous vehicles

the influence of the acceleration lane length and the visibility triangle angle of existing highway merging area on the automatic traffic flow was analyzed

and results were compared with those of traditional traffic flow. Based on the characteristics of autonomous vehicles in terms of perception

car-following

and lane-changing behaviors

as well as the collaboration with surrounding vehicles

the Krauss car-following model and LC2013 lane-changing model were improved to adapt to the characteristics of autonomous vehicles. According to the lane-changing characteristics of the vehicles

a formula was established for calculating the acceptable safety gap when changing lanes

and the car-following model parameters were adjusted while fulfilling the requirement of lane-changing safety. Results show that under the existing plane design parameters of merging area in highway

the safety

efficiency

and stability of automatic traffic flow were better than those of traditional traffic flow. Compared with traditional vehicles

the number of collisions of automatic traffic flow reduced by nearly 100%

average delay reduced by 60% to 71%

average vehicle speed increased by nearly 20%

and the speed was more stable. Under different plane design parameters of merging area

the number of collisions of automatic traffic flow was 0

the average delay was maintained at around 0.65 s

and the average speed was stable at 33-34 m/s. The existing plane design parameters of merging area in highway well adapted to the autonomous vehicles in terms of safety

efficiency

and stability

and the parameter values of the merging area had little effect on the autonomous vehicles.