Research on Lane-Change Trajectory Model for Freeway Using UAV

In order to solve the problem that the parameters of previous lane-change trajectory models lack practical meaning, this paper conducted a study on the lane change trajectory model. Firstly, the UAV was used to collect the trajectory data in a freeway diverging area, and the data was extracted by YOLOv3 target detection algorithm. Then this study analyzes the characteristics of the lane-change trajectory and the fitting of different functions to data collected. Based on the above analysis, a new S-curve lane-change trajectory model is constructed. Finally the applicability of the model was verified by goodness-of-fit (R-square) tests and simulation by the CarSim software. The study shows that the newly established lane-change trajectory model can accurately describe the lane-change trajectory, and the parameter α can reflect the degree of urgency in the process of lane-change. Based on this feature, the model is useful for research related to Advanced Driver Assistance System.