基于BP神经网络的延安市冬季PM_(2.5)浓度预测

Prediction of PM_(2.5) concentration in winter in Yan'an city based on BP neural network

提出一种基于BP神经网络的PM_(2.5)浓度预测模型,用于预测延安市冬季PM_(2.5)浓度。采用均方误差(EMS)、平均绝对误差(E_(MA))、均方根误差(E_(RMS))以及决定系数(R2)对模型的预测能力进行评估。以2021年冬季的PM_(2.5)浓度预测为例,共进行10次实验,模型的预测值和真实值之间的EMS、EMA、ERMS、R2的均值分别为29.45、4.40、5.41和0.85。研究结果表明BP神经网络模型用于预测延安市PM_(2.5)浓度是可行的。

A prediction model of PM_(2.5) concentration based on BP neural network was proposed to predict the PM_(2.5)concentration in Yan'an City.Mean square error(EMS),mean absolute error(E_(MA)),root mean square error(E_(RMS)) and coefficient of determination(R~2) were used to evaluate the predictive performance of the model.Taking the prediction of PM_(2.5) concentration in winter of 2021 as an example,ten experiments were conducted,and the mean values of EMS,EMA,ERMS and R~2 between the predicted value of model and the true value were 29.45,4.40,5.41and 0.85,respectively.It shows that BP neural network model is feasible to predict PM_(2.5) concentration in Yan 'an City.