北京大学学报自然科学版 ›› 2023, Vol. 59 ›› Issue (5): 793-800.DOI: 10.13209/j.0479-8023.2023.051

上一篇    下一篇

基于最优城市匹配神经网络模型的PM2.5插值方法

周佩1, 杨凡2,†, 韦骏1,3,†   

  1. 1. 中山大学大气科学学院, 珠海 519082 2. 国家海洋局珠海海洋环境监测中心站, 珠海 519015 3. 南方海洋科学与工程广东省实验室(珠海), 珠海 519082
  • 收稿日期:2022-09-23 修回日期:2023-04-17 出版日期:2023-09-20 发布日期:2023-09-18
  • 通讯作者: 杨凡, E-mail: yangf(at)scs.mnr.gov.cn, 韦骏, E-mail: weijun5(at)mail.sysu.edu.cn
  • 基金资助:
    广东省重点领域研发计划(2020B1111020003)和国家自然科学基金(41976007)资助

A PM2.5 Interpolation Method Based on Neural Network for Optimum City Matching

ZHOU Pei1, YANG Fan2,†, WEI Jun1,3,†   

  1. 1. School of Atmospheric Sciences, Sun Yat-Sen University, Zhuhai 519082 2. Zhuhai Marine Environmental Monitoring Central Station of the State Oceanic Administration, Zhuhai 519015 3. Southern Marine Science and Engineering Guangdong Laboratory (Zhuhai), Guangdong 519082
  • Received:2022-09-23 Revised:2023-04-17 Online:2023-09-20 Published:2023-09-18
  • Contact: YANG Fan, E-mail: yangf(at)scs.mnr.gov.cn, WEI Jun, E-mail: weijun5(at)mail.sysu.edu.cn

PDF

摘要：

为解决部分城市PM2.5浓度数据缺值严重, 无法通过训练自身数据得到预报模型的问题, 提出用相似城市的预报模型实现目标城市历史数据的填补。依据23个城市的气象数据、城市发展数据和PM2.5浓度数据, 建立基于自组织映射(SOM)和门控循环单元(GRU)神经网络的PM2.5日均浓度数据插值模型, 并分别利用该插值模型和传统插值方法(线性插值和样条插值)对不同类型的缺值数据进行填补, 对比两者的填补效果。实验结果表明, 基于SOM神经网络的城市匹配模型可以准确地匹配出目标城市的相似城市; 当缺值数据少于5天时, 利用传统插值方法的填补效果优于GRU插值模型; 当缺值数据多于5天时, GRU插值模型更胜任长时间缺测数据的填补工作。

关键词: PM2.5, 自组织映射(SOM), 门控循环单元(GRU), 插值模型

Abstract:

In order to solve the problem that some cities have serious PM2.5 concentration data deficiency and can not get prediction model by training their own data, this paper proposes a method to use prediction model of similar cities to fill in historical data of target city. Based on the meteorological data, urban development data and PM2.5 concentration data of 23 cities, an interpolation model of PM2.5 daily concentration data is established based on self-organizing map (SOM) and gated recurrent unit (GRU) neural network and use it and raditional interpolation (linear interpolation and quadratic interpolation) methods are used to fill missing data respectively. The comparision of filling effects show that the SOM city matching model can accurately match the similar cities of the target city. For the missing data less than 5 days, the filling effect of traditional interpolation methods are better than that of GRU interpolation model. For the missing data more than 5 days, the GRU interpolation model is more competent to fill in the missing data for a long time.

Key words: PM2.5, self-organizing map (SOM), gated recurrent unit (GRU), interpolation model

版权所有 © 《北京大学学报（自然科学版）》编辑部
技术支持：北京玛格泰克科技发展有限公司