Traffic volume prediction of vehicle global path planning at highway entrances and exits

This paper presents a traffic volume prediction method for global vehicle path planning at freeway entrances and exits. Firstly, a "point-line" vehicle path planning model was established, and the vehicle driving data was obtained by combining CCD sensor camera. Secondly, the weight of holidays in traffic volume prediction is analyzed from time Angle. The weight and speed in traffic volume prediction are analyzed from the perspective of space, and the weight of vehicle state index is calculated by using adaptive weighted data fusion algorithm. Finally, the traffic volume prediction algorithm is designed to realize the traffic volume prediction of vehicle global path planning. The results show that in the long-term traffic volume forecast results through this method, the Japanese and RMSE are stable at 5.44 and 9.06 respectively and the difference between the daily traffic volume forecast results and the actual results is in 10, which is obviously higher than the comparison method with high accuracy.