SA-AGA-ELM的PM_(2.5)浓度预测模型研究  

Research on PM_(2.5) Concentration Prediction Model of AGACombination SA Optimization ELM

作　　者：陈优良[1] 陶剑辉 王兆茹 陈洋[1] CHEN You-liang;TAO Jian-hui;WANG Zhao-ru;CHEN Yang(School of Civil and Surveying Engineering,Jiangxi University of Science and Technology,Ganzhou 341000,China;School of Resources and Environmental Engineering,Jiangxi University of Science and Technology,Ganzhou 341000,China)

机构地区：[1]江西理工大学土木与测绘工程学院,江西赣州341000 [2]江西理工大学资源与环境工程学院,江西赣州341000

出　　处：《数学的实践与认识》2021年第21期156-166,共11页Mathematics in Practice and Theory

基　　金：国家自然科学基金(41261093);江西省教育厅科技项目(GJJ170522)。

摘　　要：为提高PM_(2.5)浓度的预测精度,以南昌市2019年的空气质量数据和气象数据为原始样本数据,通过相关性分析确定输入PM_(2.5)浓度预测模型的特征值,同时在自适应遗传算法(AGA)的基础上融合模拟退火算法(SA),用于优化极限学习机(ELM)的网络参数,有效避免在优化过程中参数陷入局部极值,最终建立一种基于SA-AGA-ELM的PM_(2.5)浓度预测模型。实验结果表明,ELM模型预测PM_(2.5)浓度的平均绝对误差(MAE)、均方误差(MSE)、决定系数(R^(2))分别为13.644、263.935、.879,而SA-AGA-ELM模型预测结果的MAE、MSE、R^(2)分别为3.966、28.630、.952.因此,SA-AGA-ELM模型的拟合效果更好,能够更为准确的预测PM_(2.5)浓度的变化情况.In order to improve the prediction accuracy of PM_(2.5) concentration,the air quality data and meteorological data of Nanchang City in 2019 were used as the original sample data,and the characteristic value of the input PM_(2.5) concentration prediction model was determined through correlation analysis.At the same time,the adaptive genetic algorithmbased on the fusion of simula ted annealing algori thm,used to optimize the network parame ters of the extreme learning machine,effectively avoiding the parameters from falling into local extreme values during the optimization process,and finally constructing the PM_(2.5) concentration of SA-AGA-ELM Forecast model.The experimental results showed that the average absolute error,mean square error,and coefficient of determination of the PM_(2.5) concentration predicted by the ELM model are 13.644,263.935,and 0.879,respectively,while the SA-AGA-ELM model predicts the result of the average absolute error,mean square error,and coefficient of determination are 3.966,28.630 and 0.952 respectively.Therefore,the SA-AGA-ELM model has a bet ter fitting effec t and can more accura tely predic t the change of PM_(2.5) concen tration.

关 键 词：PM_(2.5)浓度 相关性分析 自适应遗传算法 模拟退火算法 极限学习机 

分 类 号：X513[环境科学与工程—环境工程] TP18[自动化与计算机技术—控制理论与控制工程]