两种基于浮动车数据的实时路况预测模型比较

Comparison of Two Real-Time Traffic Prediction Models Based on Floating Car Data

通过道路的浮动车轨迹点历史数据与交通GIS信息进行统计分析,计算出市内每条道路的路况预测模型的过程。通过大量的数据分析,最终采用了模糊聚类与支持向量机方法来对路况预测模型进行表达与存储。最后对深圳市浮动车数据进行了路况预测,预测结果表明,基于这两种方法所构建的模型可在交通路况预测中实现良好的效果。

On the basis of the statistical analysis of historical data of floating car track points and GIS information,this paper describes the process of calculating the traffic forecast model each road of inside the city. Through a large amount of data analysis,this paper eventually adopts fuzzy clustering and support vector machine（SVM）method to express and store traffic forecast model. The traffic prediction is studied by using Shenzhen floating car data,and prediction results show that the model based on the two methods can achieve good effect in the road traffic prediction.