四川环境2025，Vol.44Issue(3)：28-33,6.DOI:10.14034/j.cnki.schj.2025.03.004
基于机器学习的臭氧污染预报研究
Research on Prediction of Ozone Pollution Based on Machine Learning:A Case Study of Zhenhai District,Ningbo City
摘要
Abstract
In order to study the development of an ozone prediction model,Zhenhai District,Ningbo City was taken as a case study.Based on the pollutant monitoring data,meteorological monitoring data,potential factors of ozone pollution and other forecasting variables,key forecasting factors for modeling were screened,the data was randomly divided into a training set and a test set,and nine different models were applied by using the machine learning method.We respectively established the ozone pollution forecasting model of Zhenhai District and test it,and screen the most effective model for ozone forecasting in Zhenhai District.The data from 2018-2022 were applied as a training set to model the ozone forecasting in Zhenhai district with the best model screened,and the data from 2023 were used as the test set to carry out independent tests.The results showed that the asymptotic gradient regression tree model has the best effect on ozone forecasting in Zhenhai District,and the ozone forecasting model constructed by applying it has a better forecasting ability in predicting the ozone concentration level.In the daily ozone maximum 8-hour average concentration forecast for 2023,the accuracy rate of ozone pollution level prediction was 72.0％,and the absolute error of the daily ozone maximal 8-hour average concentration forecast was 19.24 g/m3.The ozone forecasting model has good application value in ozone forecasting in Zhenhai District.关键词
臭氧/机器学习/预报因子Key words
Ozone/machine learning/forecasting factor分类
环境科学引用本文复制引用
申子彬,庞乐,王文典,郑艳,张国超..基于机器学习的臭氧污染预报研究[J].四川环境,2025,44(3):28-33,6.基金项目
镇海区公益类科技计划项目(镇科[2022]23号-15). （镇科[2022]23号-15）
