A New Algorithm for Solving the Parameters of Approximate Non Homogeneous Exponential Sequence Model

In response to the application scope of the model being an approximate non-homogeneous exponential sequence, first, a simple method for solving the developmental quotient coefficient of inhomogeneous exponential Grey prediction formula is given.After the developmental quotient is determined, then the prediction coefficient and shift constant can be worked out based on the known value of the first and last item, thus the developmental quotient can thus be directly estimated.Such a method, can either be combined with the method of determining the predictive coefficient and translation constant through the use of the first and last item as the initial condition, thus to develop a brand new model, or be joined with the method of determining the predictive coefficient and translation constant by minimizing the relative error via averaging as the goal to achieve a new improved model.The theoretical and practical results show that the new model can improve the modeling accuracy to a certain extent.This method works well for both the traditional model and the new model.