GM(1 ,1) model of time sequence coefficient and application

Based on the error analysis of the GM (1, 1) model, this paper corrects the equidistance sequence and establishes GM (1 ,1) model of time sequence coefficient with time coefficient. This model overcomes the disadvantage which the residual sequence obtained by residual model may not satisfy the condition of no negativity. The new method simplifies the modeling process and has higher precision of forecast and simulation. Therefore this model had certain theory value and the application value Example shows that GM (1, 1) model of time sequence coefficient is more suitable in fluctuant primary data and also indicates its usability and the reliability.