基于轨迹数据的快速路交织区拥堵演变特征研究

Research on the Congestion Evolution Characteristics of Expressway Weaving Area Based on Trajectory Data

对快速路交织区拥堵演变过程中宏观交通流参数与交通状态的时变关系进行研究,可以为快速路交织区交通状态判别提供科学依据。利用YOLO算法从高清视频中提取车辆轨迹数据后,利用卡尔曼滤波对原始轨迹数据进行降噪平滑处理;对快速路交织区拥堵演变过程中速度、流量、密度等宏观交通流参数与交通状态进行时变分析,揭示快速路交织区宏观交通流参数在拥堵演变过程中的时变特征。结果表明,在快速路交织区交通状态判别时,融合平均行程速度和交通流密度等指标,可以有效提高交通状态判别精度。

Studying the time-varying relationship between macro traffic flow parameters and traffic state during the evolution of congestion can provide scientific basis for the identification of traffic state in the weaving area of expressway.After extracting vehicle trajectory data from high-definition video by YOLO algorithm,the original trajectory data is denoised and smoothed by Kalman filter.This paper analyzes the time-varying characteristics of macro traffic flow parameters such as speed,flow and density and traffic state in the congestion evolution of expressway weaving area,and reveals the time-varying characteristics of macro traffic flow parameters in the congestion evolution of expressway weaving area.The results indicate that integrating indicators such as average travel speed and traffic flow density can effectively improve the accuracy of traffic state discrimination in the weaving area of expressway.