Inclement weather conditions may have a significant influence on traffic state estimation and prediction. To achieve more accurate estimation and prediction, it is essential to consider meteorological influence in the traffic flow estimation. This study first introduces an image-based method to analyze relevant meteorological features, using meteorological images data from the China National Meteorological Administration. We extract the features of hourly rainfall, hourly temperature, hourly humidity, and daily wind speed data across the country. The relationship between variables in the traffic flow model in various weather conditions is analysis in Ningbo City by making using of the traffic survey station data. The experiment results show that temperature, illumination, and precipitation have a greater impact on some parameters in the traffic flow model, such as flow rate and average speed. Therefore, the traffic flow model analysis incorporating weather impacts can be effectively used for traffic state estimation and prediction under rainy conditions in Ningbo City.

CICTP

暂无