文章基本信息

标题：Evolving Spiking Neural Network Model for PM2.5Hourly Concentration Prediction Based on SeasonalDifferences: A Case Study on Data from Beijing andShanghai
本地全文：下载
作者：Hengyuan Liu ; Guibin Lu ; Yangjun Wang 等
期刊名称：Aerosol and Air Quality Research
印刷版ISSN：1680-8584
出版年度：2021
卷号：21
期号：2
页码：1-15
DOI：10.4209/aaqr.2020.05.0247
语种：English
出版社：Chinese Association for Aerosol Research in Taiwan
摘要：In recent years, the dangers that air pollutants pose to human health and the environmenthave received widespread attention. Although accurately predicting the air quality is essential tomanaging pollution and developing control policies, traditional forecasting models have not beenable to simulate the seasonal and diurnal variation in air pollutant concentrations. Furthermore,inadequate processing of the available spatio-temporal data has precluded the capture ofpredictive historical patterns. Therefore, we have developed a staging evolving spiking neuralnetwork (eSNN) model named Staging-eSNN that first employs a time series clustering algorithmto distinguish the seasonal from the diurnal variation in the PM2.5 concentration. We then predictthe concentrationsin Beijing and Shanghai 1, 3, 6, 12 and 24 hours in advance. Various evaluationindicators show that the Staging-eSNN model achieves higher performance than the supportvector regression (SVR), random forest (RF) and other eSNN models.
关键词：Air pollutant prediction;PM25 hourly concentration;Seasonality;Evolving spikingneural networks;Time series clustering