论文部分内容阅读
基于车辆速度参数是描述交通流运行的最基本参数之一,在交通状态评价与交通安全性分析中具有的重要作用,针对城市道路不同种类车型混合运行的特征,提出采用高斯混合模型来描述自由流状态下的车辆速度分布。结合实测速度数据中的双峰分布特性,将混合车流车型分为大型车与小型车2类,进而建立了2类型的高斯混合模型,并采用期望最大化算法对模型参数进行估计。通过实测数据验证结果表明,模型能够很好的拟合观测速度数据,并能够有效区分不同类型车辆速度,为确定不同类型车辆限速标准、分析不同类型车辆相互影响提供理论依据。
Based on the fact that vehicle speed parameter is one of the most basic parameters to describe traffic flow operation and plays an important role in traffic status evaluation and traffic safety analysis, aiming at the characteristics of mixed operation of different types of vehicles on urban roads, a Gaussian mixture model is proposed to describe the freedom Vehicle speed distribution in flow state. Combining with the bimodal distribution in the measured velocity data, the hybrid vehicle model is divided into two types: large vehicle and small vehicle. Then two kinds of Gaussian mixture models are established, and the model of parameter estimation is estimated by the expectation maximization algorithm. The experimental results show that the model can well fit the observed velocity data, and can effectively distinguish the different types of vehicle speed, in order to determine the different types of vehicle speed limit standards, analyze the interaction between different types of vehicles to provide a theoretical basis.