多元线性回归与灰色联合模型在湖泊水质预测中的应用

Application of the Combined method of Multivariate Linearity Regression and Grey Model in Forecasting the Water Quality

根据对湖泊水体中所含高锰酸盐指数、氨氮、总磷等主要化学指标及其相关因素人口、工业产值和污水排放量的数据,采用了多元线性回归方法建立了湖泊水质预测的数学模型。并用多元线性回归和灰色模型联合的方法进行实际预测,多元线性回归模型通过F检验认为回归方程的效果比较显著,灰色模型精度检验达到二级,收到了良好效果。

According to the primary chemistry indexes （such as the index of the salts of permanganic acid, NH3-N and TP） in a lake and the related factors （such as population, industry production value and sewage quantity）, by using the method of multivariate linearity regression, the mathematical model in predicting the water quality was established. The predicted value could be calculated by a combination of multivariate linearity regression and Grey Model. The effect of multivariate linearity regression by F test could be more accurate and the precision checkout of Grey Model could be achieved.