一种基于随机配置网络的PM2.5浓度预测方法

A Method of Predicting PM2.5 Using Stochastic Configuration Network

作为评价空气质量的重要指标,细颗粒物(Fine Particulate Matter, PM2.5)浓度是实现环境治理的基础。降低PM2.5浓度可有效改善空气质量,减少各种呼吸道病况的发生。因此,PM2.5浓度的预测变得尤为重要。本文基于随机配置网络算法,建立一个非线性回归模型用于预测PM2.5的浓度。实验结果表明:采用随机配置网络算法建立的PM2.5浓度预测模型具有较高的预测精度。

As a key index of measuring the air quality, fine particulate matter (PM2.5) plays an important role in realizing the environmental treatment. The air quality can be effectively improved by reducing the PM2.5 concentration, which can prevent the occurrence of various kinds of respiratory diseases. Hence, it is very important to predict the PM2.5. In this paper, using the stochastic configuration network algorithm, a nonlinear regression model is established to predict the PM2.5. Experiment result illustrates that the established PM2.5 prediction model using the stochastic configuration network has a high precision.