基于MEA＿SVM空气质量指数预测  

Prediction of Air Quality Index Based on MEA-SVM

作　　者：俆乔王 胡红萍[1] 白艳萍[1] 王建中[1] XU Qiaowang;HU Hongping;BAI Yanping;WANG Jianzhong(School of Science,North University of China,Taiyuan 030051,China)

机构地区：[1]中北大学理学院

出　　处：《重庆理工大学学报（自然科学）》2019年第12期150-155,共6页Journal of Chongqing University of Technology：Natural Science

基　　金：国家自然科学基金项目(61774137);山西省自然科学基金项目(201801D121026,201701D121012,201701D221121);山西省回国留学人员科研项目(2016-088)

摘　　要：针对大气污染物的动态性和高时空变异性,分析了影响太原市空气质量评价的主要污染物,建立思维进化算法(MEA)与支持向量机(SVM)结合的新模型MEA_SVM,并将该模型用于空气质量指数(AQI)的预测。实验结果表明:MEA_SVM算法在保证SVM预测准确率的同时显著提高了预测速度,在预测可靠性、预测精度方面均优于遗传算法与SVM的结合方法和粒子群算法与SVM结合的方法,因此MEA_SVM算法在城市空气质量预测中有一定的实用价值。In view of the dynamic nature and high temporal and spatial variability of air pollutants,this paper analyzes the main pollutants affecting air quality evaluation in Taiyuan city,establishes MEA_SVM,a new model combining mind evolutionary algorithm(MEA)and support vector machine(SVM),and applies this model to the prediction of air quality index(AQI).The experimental results show that the MEA_SVM algorithm not only guarantees the accuracy of SVM prediction,but also significantly improves the prediction speed.In terms of prediction reliability and accuracy,it is superior to the combination of genetic algorithm with SVM and particle swarm optimization with SVM.Therefore,MEA_SVM algorithm has certain practical value in urban air quality prediction.

关 键 词：思维进化算法 遗传算法 粒子群算法 支持向量机 信息粒化 空气质量指数预测 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]