车联网环境下的动态Robertson车队离散模型

传统Robertson车队离散模型参数估计是基于历史数据，不能很好地反映交通流的动态变化特征，为解决这一问题，构建了车联网环境下的动态Robertson车队离散模型.考虑到车联网环境下车辆的行程时间数据易于获得，基于此可对Robertson模型的相关参数进行实时动态估计建立动态Robertson车队流量离散模型.通过实际调查数据，分析了上游交叉口车辆离去流率与下游交叉口车辆到达流率的关系，并将文中模型与静态Robertson模型、实际观测数据进行了比较分析.结果表明，文中动态模型更能反映交通流的车队离散规律，与静态Robertson模型相比，平均预测均方误差减少了30.68%. 

Dynamic Robertson's Platoon Dispersion Model in Connected Vehicle Environment

The parameters of the traditional Robertson's platoon dispersion model are based on historical data, and thus, it cannot provide a good reflection of the dynamic characteristics of traffic flow. In order to solve this problem, a connected vehicle environment is considered, and the travel time of the vehicles is easily obtained. The relevant parameters of the Robertson's model can be estimated in real time based on this. Then, a dynamic Robertson's platoon dispersion model was proposed. Later, the relationship between the arrival flow rate of the downstream intersection and the departing flow rate of the upstream intersection was analysed using the proposed model with field collected data, and compared with those obtained using the traditional Robertson's model and actual data. The results show that the proposed model can better describe the law of dispersion in traffic flow, and the mean squared error of prediction is reduced by approximately 30.68%, compared with the traditional Robertson's model.