Comparison of grain yield prediction methods in land use planning

In order to explore the ways of increasing grain yield prediction accuracy in land use planning, related data of grain yield was analyzed between 1988 to 2005 in Jincheng, and the linear regression model, gray GM (1,1) model and gray- multiple linear regression model was compared. Firstly, using grey incidence analysis, factors affecting the grain yield were sorted. Secondly, based on the grey incidence analysis, main factors were selected. Thirdly, through gray GM (1,1) model, the predictive value of main factors was calculated. At the same time, the multiple linear regression model was also constructed using original data. The last, taking predictions of gray GM (1,1) model as input value, gray- multiple linear regression model was constructed. The result showed that gray- multiple linear regression model had higher prediction accuracy than linear regression model and gray GM (1,1) model ,it was the most appropriate model to forecast the grain yield in Jincheng. The research could improve the scientific of establishment in land use planning.