城市空气污染数据统计分析及治理措施研究

Statistical Analysis of Urban Air Pollution Data and Research on Treatment Measures

城市空气污染数据具有统计特性,根据统计结果可以实现对空气污染数据的有效预测,提出一种基于支持向量机学习的城市空气污染数据预测方法。采用多维传感器进行城市空气污染数据采集,采用决策数据高维映射重构模型对多传感器采样的城市空气污染数据先验特征进行信息重排,采用闭频繁项集挖掘方法提取城市空气污染数据的统计特征量,采用量化回归分析方法进行大数据的主特征分析,将量化回归分析结果输入到支持向量机预测器中,结合灰度学习进行污染数据的有效预测和加权控制,提高数据预测的自适应性,实现城市空气污染数据优化预测。

Urban air pollution data have statistical characteristics. Based on the statistical results, an effective prediction method of urban air pollution data based on support vector machine learning is proposed. The multi-dimensional sensor is used to collect urban air pollution data, and the decision data high-dimensional mapping reconstruction model is used to rearrange the priori characteristics of multi-sensor sampling urban air pollution data. The closed frequent item set mining method is used to extract the statistical features of urban air pollution data, and the quantitative regression analysis method is used to analyze big data s main features. The results of quantitative regression analysis are input into the support vector machine(SVM)predictor. In order to improve the self-adaptability of data prediction, the optimal prediction of urban air pollution data can be realized by combining gray learning with effective prediction and weighted control of pollution data.