Development of A Traffic Forecast System Based on LOM and Proposed Optimal Routing Method with PSO

近年来，汽车社会的发展所产生的交通拥堵、交通事故、环境污染等问题已经成为一个严重的社会问题。因此，交通预测和分析对于减少交通拥堵和改善环境是不可或缺的。此外，利用路侧传感器的实时信息进行高效、安全和流畅的交通控制技术也是人们所期待的。利用实时交通数据，采用基于模型的预测方法，对各观测点30分钟内的交通密度进行了估算研究。提出了一种基于LOM和平滑滤波的交通量预测系统。此外，通过对PSO计算的估计交通密度和旅行距离进行综合评估，可以实时地给出从当前地点到目的地的最优路径。

In recent years, traffic jams, traffic accidents, environmental pollutions and etc. generated by the growth of automobile society have been become a serious social problem. Therefore, the traffic forecasting and analysis are indispensable for the reduction of traffic jams and the improvement of environment. Additionally, an efficient, safety and smooth control technology in traffic utilizing the real-time information from roadside sensors is expected. We conduct a research on estimating the traffic density in 30 minutes at every observation point by model-based prediction approach with real-time traffic data. A traffic forecasting system based on LOM with a smoothing filter is proposed. Additionally, an optimal routing form present place to destination can be presented in real time with a comprehensive evaluation of the estimated traffic densities and travel distances calculated by PSO.