基于Optima的实时在线交通流预测方法研究

目前城市道路交通流预测主要是基于物理模型、数理统计特性并融合部分智能预测算法来实现的,而对于一些影响预测效果的重要交通因素,诸如FIFO原则、交通信号控制方案等在现有的预测方法因无法很好地引入和描述而忽略.本文提出了一种基于Optima系统实现的实时在线交通预测方法,通过建立路网模型、需求模型及初始OD矩阵获取路网实时状况,并通过构建关联数据库实现实时路网模型信息、交通信号控制信息的有效对接,依据TRE算法预测路段进出口的累积流量并结合模型分配值、历史数据实现实时在线交通预测.以北京市望京区域为例进行仿真验证,通过误差分析,获得了较为理想的预测效果,验证了该预测方案的有效性.

Real-time and Online Traffic Flow Prediction Based on Optima Method

Urban road traffic flow forecast is mainly based on physical modeling, statistics features and fusion part of intelligent prediction algorithm to achieve. Important factors that influence results such as the FIFO principle, traffic signal control are always neglected due to inability to introduce and describe. This paper proposes a real- time and online traffic flow forecast method based on Optima. Through the establishment of network models, demand models and initial OD matrix gets real-time road network conditions. And by building a database model of real-time network information, traffic signal control match the information effectively. According to TRE algorithm predicting road accumulated flow, completing the real-time and online traffic flow forecast combined with the model assignment value and historical data. Taking Wangjing district in Beijing as a simulating example. Through error analysis, we obtain ideal results, and verify the effectiveness of the proposed method.