基于多维城市情报的空气质量指数协同预测研究:以京津石为例

Collaborative Forecast of Air Quality Index Based on Multidimensional City Information: a Case Study of Beijing-Tianjin-Shijiazhuang

京津冀首都经济圈空气质量的提升和改善是实现京津冀协同发展的重要问题。论文选取了北京、天津和石家庄3个城市为研究对象,分别应用不同城市空气质量情报进行空气质量指数协同预测实验。实验分成3个部分来实现,首先应用目标城市的一维城市情报进行预测,其次分别选取目标城市和另一城市的两维城市情报进行预测,最后选用3个城市的三维城市情报进行预测。实验结果表明,大部分城市通过多维城市情报的空气质量指数预测确实可以有效地提高预测的准确度,但是部分城市由于存在较强的空气特征属性协同预测效果不佳。同时通过城市产业定位分析,也得出北京、天津和石家庄3个城市的空气质量较差主要与石家庄的空气污染较严重的重工业较多有关系,可以通过产业升级和调整进行改善的结论。

The improvement of air quality was an important issue to achieve collaborative development of Beijing, Tianjin and Hebei Province. Beijing, Tianjin and Shijiazhuang were selected as study objects in this paper. The collaborative AQI forecasting experiments were carried out using multi-dimensional city information. Experiment was divided into three parts, firstly one-dimensional city information was applied to predict AQI of the target city; secondly, the two--dimensional city in- formation of target city and another city were applied to the prediction, and finally, the three-dimensional city information were selected to predict AQI of the target city. The results of experiments showed that the accuracy of AQI forecasting us- ing multi-dimensional city information mostly could be improved. But the AQI forecasting accuracy of some city was not well, because of their strong synergy property. At the same time, it could be concluded that there would be some relations be- tween the poor air quality of three cities and the heavy industry in Shijiazhuang. The poor air quality of the three cities should be improved by the industrial upgrading and adjustment.