基于熵和条件生成对抗网络的城市AQI预测研究  

Urban AQI prediction based on entropy and conditional generation countermeasure network

作　　者：文琴 何文孝[1] WEN Qin;HE Wenxiao(College of Artificial Intelligence,Neijiang Normal University,Neijiang,Sichuan 641100,China)

机构地区：[1]内江师范学院人工智能学院,四川内江641100

出　　处：《内江师范学院学报》2022年第12期49-55,62,共8页Journal of Neijiang Normal University

基　　金：内江师范学院校级科研项目(2020YB34)。

摘　　要：精准的短期城市空气质量指数(AQI)预测对人们合理规划出行、政府精确开展治理环境问题、提升城市形象、促进经济发展有着十分重要的意义.AQI的变化与气象条件和污染物的排放量等诸多因素有关,因此,城市AQI预测的研究和评价是一个多变量、非线性问题.为了深度学习非线性AQI数据中所隐含的深层次关系和提高预测的准确性,提出了一种结合气象因子多模式集成预报的基于条件生成对抗网络(CGAN)的AQI预测模型.以四川省内江市3年的AQI数据作为实际算例,对比其他模型的预测结果,仿真实验结果表明该预测模型对AQI的预测结果更加准确.Accurate short-term urban air quality index(AQI) prediction is of great significance for people to reasonably plan their own travel, and for the government to accurately deal with environmental problems, thus to improve the city’s image and promote economic development. The change of AQI is related to many factors such as meteorological conditions and pollutant emissions. Therefore, the research and evaluation of urban AQI prediction is a multi-variable and nonlinear problem. In order to deeply learn the deep-seated relationship implied in nonlinear AQI data and improve the accuracy of prediction, an AQI prediction model based on conditional generation countermeasure network(CGAN) combined with multi-mode integrated prediction of meteorological factors is proposed. Taking the AQI data of Neijiang City in Sichuan Province for three years as an example, compared with the prediction results of other models, the simulation results show that the prediction model is more accurate in predicting AQI.

关 键 词：条件生成对抗网络 AQI 多模式集成预报 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术]