Research on Grain Yield Forecasting System Based on Grey Model

The grain yield forecasting system is an important part of the food security, which crucially related with the safety of the politics and the military of the country, however, existing prediction methods have reduced accuracy due to problems such as data correlation and data sparsity. To overcome these problems, the grey forecast (GF) model for the grain yield is proposed in this paper. Firstly, the forecasting method based on the grey theory is established by using the yield series data from 1980 to 2012. Then, the prediction model is tested by the residual, the posteriori difference and the correlation coefficient. Lastly, the prediction precision is compared between the GM(1,1) and the moving average methods. The results shown that the forecasting model can predict the grain yield accurately, the mean relative error is 3.78%, and the maximum error is less than 10%, which can satisfy the needs of the forecasting system.