基于气象因素和改进支持向量机的空气质量指数预测

空气污染已成为人类社会面临的严峻挑战,为了构建准确的空气质量预测模型,文章首先运用统计分析方法进行相关性分析,探讨气象要素变化对空气质量的影响;针对传统支持向量机预测精确度受输入变量影响较大的弊端,采用基于熵权理论对变精度粗糙集约简进行了改进,以处理支持向量机的输入变量,提高支持向量机的预测精度;最后以沈阳市的历史气象数据和空气质量指数进行验证。与传统方法相比,所提方法能够将预测准确率由71.28%提高到77.83%,空报率和漏报率也有降低,与实际基本吻合。

AIR QUALITY INDEX ESTIMATION METHOD BASED ON METEOROLOGICAL ELEMENTS DATA AND MODIFIED SUPPORT VECTOR MACHINE

Air pollution has become a serious challenge to human society,in order to construct an accurate air quality index prediction model,this paper first analyzed the impact of meteorological factors on air quality index by using Spearman rank correlation coefficient.In view of the traditional support vector machine (SVM)forecasting precision were greatly influenced by the input variable,variable precision rough set (VPRS) was modified based on entropy weight theory and the prediction accuracy of support vector machine improved.Finally,the historical meteorological data and air quality index of Shenyang were taken as the basis for verification.The results showed that the proposed method could improve the prediction accuracy from 71.28％ to 77.83％ compared with the traditional method,and the rate of false report and false negative rate also decreased obviously,which verified the effectiveness of the proposed method.