Characteristics of Negative Oxygen Ion Concentration and Its Prediction Model in Tianjin
Using observation data of the negative oxygen ion(NOI)concentration,meteorological and environmental elements from June 2019 to December 2022 in Liangzhuangzi station(forest zone)and urban station(residential zone),the change characteristics of daily NOI concentration in different time scales and its relationship with different factors in Tianjin are analyzed.The best prediction model of NOI concentration based on machine learning is built.The results show that:(1)The NOI concentration in forest and urban zones shows"single peak-single valley"pattern in daily scale,and in forest zone,there is"double peak-single valley"pattern in monthly scale,but in urban zone,there is no significant pattern.(2)The NOI concentration in forest zone is ranked as spring and winter＞autumn＞summer,and that in urban zone is ranked as summer＞spring＞autumn＞winter.(3)The NOI concentration in urban zone is lower than that in forest zone,and there are different influencing factors in different zones.(4)The random forest method is more suitable for building the prediction model of NOI concentration in forest zone,and the stochastic gradient descent method is more suitable for urban zone.
negative oxygen ion concentrationmeteorological elementsprediction modelmachine learning
万方数据
维普
