多城市空气质量指数的分析与预测

发布时间：2018-06-20 11:59

本文选题：空气质量指数 + 统计分析　；参考：《兰州大学》2017年硕士论文

【摘要】：随着我国大部分地区雾霾天气的频繁出现,环境问题已经引起人们的普遍关注。如何有效的防控日益严峻的空气污染所带来的危害和影响,是当今社会亟待解决的热点问题。对空气质量进行有效的监测、预报和控制是改善大气质量的核心手段。本文主要对北京、上海、广州、兰州四个代表性城市的空气质量指数(AQI)进行相关的统计分析,并对其未来五天内空气质量变化趋势的预测研究,在针对大气污染的防控工作上有一定的实际应用价值。首先文中对四个城市一年的日AQI进行了统计分析。研究发现四个城市的日AQI变化存在很大的差异。特别地,北京市的空气质量指数变化幅度最大,四级以上的污染天气最多;广州市整体空气质量较好,四级以上污染天气最少;同一城市冬季的空气质量较差,高污染天气较多。针对这些研究结果,文中提出了几点建议。为了能够做好空气质量的预报工作,本文对四个城市实时AQI根据季节的划分分别进行了预测研究。首先利用EMD分解技术以及归一化原理对不同的数据序列进行了预处理;然后利用基于人工智能算法优化的SVM模型对各城市空气质量指数的季节性进行了五天的多步预测;最后利用误差评定标准对各模型的预测性能进行了比较。研究表明:对于不同城市的AQI季节性的预测研究并不存在一个最优的、通用的模型,但经过优化的SVM模型预测精度高于SVM模型的预测精度,且预测模型的一步预测优于多步预测的效果。
[Abstract]:With the frequent occurrence of smog weather in most areas of China, environmental problems have attracted widespread attention. How to effectively prevent and control the increasingly severe air pollution caused by the harm and impact, is a hot issue to be solved. Effective air quality monitoring, prediction and control are the core means to improve air quality. This paper mainly analyzes the air quality index (AQI) of four representative cities in Beijing, Shanghai, Guangzhou and Lanzhou, and forecasts the trend of air quality change in the next five days. It has some practical application value in the prevention and control of air pollution. Firstly, the paper makes a statistical analysis on the daily AQI of four cities in one year. The study found that there are great differences in daily AQI changes in four cities. In particular, Beijing has the largest change in the air quality index, with the most polluted weather over four levels; Guangzhou has a better overall air quality and the least polluted weather above grade four; the air quality of the same city is poor in winter. High pollution weather is more frequent. In view of these research results, some suggestions are put forward in this paper. In order to forecast the air quality, the real time AQI of four cities is predicted according to the seasonal division. Firstly, different data sequences are preprocessed by using EMD decomposition technique and normalization principle, and then the seasonal air quality index of cities is predicted by SVM model based on artificial intelligence algorithm in five days. Finally, the prediction performance of each model is compared by using error evaluation standard. The results show that there is not an optimal and universal model for the seasonal prediction of AQI in different cities, but the prediction accuracy of the optimized SVM model is higher than that of the SVM model. The one-step prediction of the prediction model is better than that of multi-step prediction.
【学位授予单位】：兰州大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：X51

【相似文献】