The Research And Application Of Grey Forecast Model

The grey system theory aim to establish a grey differential prediction model through the small amount of incomplete information, it can make a long-term description for the development rules of the complex fuzziness system, its research focus on the uncertainty problem with "poor information" and "less data". Grey prediction not only is an important part of the grey system theory, but also be widely used in predictive method. Therefore, it has great significance for research on Grey Forecast Model.Based on a large of research on grey system theory, this paper aims to improve the inadequacies which still exist in the model. First, through analyzing disadvantages of traditional modeling methods, constructs a new background value formula to improve the prediction accuracy of the model from the geometric meaning of the model background value, and then gives a new model which called IGM(1,1). Secondly, this paper takes advantage of genetic algorithm to search the global optimal solution to further improve the solving method of IGM (1,1) model’s parameters, GAGM (1,1) model based on genetic algorithm theory is proposed. Eventually, the three models are applied to predict grain yield in Inner Mongolia, by contrast with the traditional GM (1,1) model, to verify the validity and applicability of the new model. Data simulation results show that the development coefficient of improved model has little effect on simulation and prediction accuracy of this model, and it breaks the limit in traditional GM (1,1) model, the traditional GM (1,1) model’s prediction accuracy has been greatly improved.