Based on simpson formula improved non-interval GM(1,1) model and application

Non-interval gray GM (1,1) model is a tradition from the GM (1,1) model ranging parameters. Traditional unequal interval gray GM (1,1) model prediction accuracy is lower, in order to improve the unequal interval gray GM (1,1) model prediction accuracy, through the establishment of gray model based on the new rates, and analyze the background value sources of error; then using Simpson formula to improve the background value, By predicting that: non-interval improved gray GM (1,1) model accuracy is higher and apply to a summation was the number of regular sequence; their average error is less than 0.1%, compared with the traditional interval gray GM (1,1) model 0.35%; the over fitting polynomial 4.23%. Therefore, the unequal interval gray GM (1,1) model can be improved to meet the accuracy requirements of the higher forecast instances.