自建点空气质量数据校准的回归模型

Regression Model of Air Quality Data Calibration at Self Built Sites

空气质量检测仪采集的数据存在误差,需要基于国控点数据对其进行校准。首先对国控点和自建点的“两尘四气”数据进行探索性分析,包括数据预处理、时间匹配设置、差异性分析、相关性分析等。其次分别建立一元线性回归模型、多元回归模型,对自建点数据进行校准,并对模型进行检验、对比和误差分析。结果表明:多元回归模型整体更优,校准结果更为准确,可以提高检测仪的精度,科学反映实时空气质量。

The data collected by the air quality detector has errors,so it needs to be calibrated based on the data of national control points.Firstly,exploratory analysis is made on the“two dusts and four gases”data of national control points and self built points,including data preprocessing,time matching setting,difference analysis,correlation analysis,etc.Secondly,the univariate linear regression model and multivariate regression model are established to calibrate the self built point data,and the models are tested,compared and analyzed.The results show that the multiple regression model is better as a whole and the calibration results are more accurate.It can improve the accuracy of the detector and scientifically reflect the real-time air quality.