基于MATLAB和SPSS的空气质量数据建模及分析校准

利用Excel对空气质量国控点与自建点的"两尘四气"浓度日平均数据进行透视分析,得到二者相对应的变化趋势,发现大部分数据波动比较大;再利用SPSS对"两尘四气"数据与风速、压强、降水量、温度、湿度的相关性进行分析,结果表明,PM2.5、PM10与湿度呈中度相关,SO₂与风速、湿度基本不相关;最后以自建点整点时刻"两尘四气"的平均数据与国控点数据的差值为因变量,风速、压强、降水量、温度、湿度为自变量,用MATLAB进行多元线性回归,建立自建点数据的矫正模型,模型校验效果达到预期。

Mathematical Model and Calibration for Data of Air Quality Based on MATLAB and SPSS

A perspective analysis is conducted on the daily average data of"two dusts and four gases"of air quality at both national control points and self-built points by using Excel.The corresponding trend of change is obtained.It is found that most of the data fluctuate greatly.Then a second analysis is done on the correlation between the data of"two dusts and four gases"and wind speed,pressure,precipitation,temperature and humidity by using SPSS.It shows that PM2.5 and PM10 are moderately correlated with humidity,while SO₂is basically not correlated with wind speed and humidity.Finally,taking the difference between the average data of"two dust and four gas"and the national control point data as the dependent variable,and the wind speed,pressure,precipitation,temperature and humidity as the independent variable,a correction model for self-built point data is created by using MATLAB to do multiple linear regression,and the results show that the verification of model reaches the expectation.