Scholarlink

Ozone (O₃) and fine particulate matter (PM2.5) are two pollutants which have the greatest impact on the urban atmospheric environment. In recent years, O₃ outranked PM2.5 as the most important air pollutant in some cities, bringing major hidden dangers to human health. Prediction of air pollution becomes a hot topic of all fields. In this paper, a variable Wavelet-ARIMA (VWA) forecasting model is proposed. The O₃ dataset in 2020 is classified into three groups by numerical characteristics, which directly derive the different parameters of wavelet functions. Then, Autoregressive Integral Moving Average (ARIMA) model is applied on wavelet reconstructed sequences to produce three parts of predictions. By mingling these parts, we get the total ozone concentrations for the next 5 days, 7 days and 10 days. Finally, the superiority on new model is illustrated over the traditional ones. The revised model is superior to the traditional one, and provides a new idea for pollutant prediction.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A Revised Method Based on Wavelet-Time Series and the Application in Prediction of Ozone Concentration

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Yuchen Wu,
  • Chenguang Yan,
  • Xiaoyi Chen

摘要

Ozone (O₃) and fine particulate matter (PM2.5) are two pollutants which have the greatest impact on the urban atmospheric environment. In recent years, O₃ outranked PM2.5 as the most important air pollutant in some cities, bringing major hidden dangers to human health. Prediction of air pollution becomes a hot topic of all fields. In this paper, a variable Wavelet-ARIMA (VWA) forecasting model is proposed. The O₃ dataset in 2020 is classified into three groups by numerical characteristics, which directly derive the different parameters of wavelet functions. Then, Autoregressive Integral Moving Average (ARIMA) model is applied on wavelet reconstructed sequences to produce three parts of predictions. By mingling these parts, we get the total ozone concentrations for the next 5 days, 7 days and 10 days. Finally, the superiority on new model is illustrated over the traditional ones. The revised model is superior to the traditional one, and provides a new idea for pollutant prediction.