关键词: 玉米产量, 多元线性回归, 灰色模型, 日照市

Abstract: Based on the corn yield data and related meteorological data of Rizhao City,the prediction model of corn yield was constructed by stepwise regression and grey theory. Using MATLAB software to establish a multivariate linear regression equation,the equation is predicted by the significant level of 0.05, the predicted yield error of 1998-2008 is 0.5%~3.6%,and the prediction accuracy is high; using gray system theory, GM（1,N） The grey prediction model is improved. The predicted output error range from 1994 to 2006 is 1.4%~16.6%,and the predicted accuracy is about 90%. The effect is better.

Key words: maize yield, multiple linear regression, GM model, Rizhao city