The invention relates to a vehicle track prediction method for a detector-free road section. The method comprises the following steps: S1, collecting real vehicle historical data of a target detector-free road section; s2, initializing training parameters of an XGBoost model for predicting driving-out time and an XGBoost model for predicting a target lane according to historical data of a real vehicle; s3, inputting training parameters into the two models for training, enabling the XGBoost model for predicting the driving-out time to output the predicted driving-out time of the virtual vehicle, and enabling the XGBoost model for predicting the target lane to output the predicted target lane of the virtual vehicle; s4, inputting actual parameters, respectively obtaining predicted driving-out time and a predicted target lane according to the two different models, and calculating the driving duration of the virtual vehicle; s5, calculating the speed according to the total distance and the driving time of the target detector-free road section; and S6, controlling the virtual vehicle to run according to the calculated speed and the predicted target lane. According to the invention, high-accuracy and easy-to-implement vehicle trajectory prediction can be realized in a detector-free road section.
本发明涉及一种无检测器路段的车辆轨迹预测方法,包括以下步骤:S1:收集目标无检测器路段的真车历史数据;S2:根据真车历史数据初始化预测驶出时间的XGBoost模型和预测目标车道的XGBoost模型的训练参数;S3:将训练参数输入两个模型进行训练,使预测驶出时间的XGBoost模型输出虚拟车辆的预测驶出时间,使预测目标车道的XGBoost模型输出虚拟车辆的预测目标车道;S4:输入实际参数,根据两个不同的模型分别得预测驶出时间和预测目标车道,并计算虚拟车辆的行驶时长;S5:根据目标无检测器路段总路程及行驶时长计算得到速度;S6:控制虚拟车辆按照计算的速度及预测的目标车道行驶。本发明可以在无检测器路段实现准确率高且容易实施的车辆轨迹预测。
Vehicle track prediction method and system for road section without detector
一种无检测器路段的车辆轨迹预测方法及系统
2023-09-12
Patent
Electronic Resource
Chinese
Smooth transition method for track from vehicle intersection to road section
European Patent Office | 2024
|European Patent Office | 2021
|Road section average vehicle speed calculation method based on floating vehicle virtual track points
European Patent Office | 2024
|