An Improved GM(1,1) Model of Integrated Optimizing Its Background Value and Initial Condition

摘要

Although GM(1,1) model has been successfully adopted in various fields and demonstrated promising results, its predicting performance still could be improved. Up to the present, the literatures show that whether the structure method of background value or the selection of initial condition is logical or not, it affects the simulation and prediction precision directly. However, most optimized models developed recently are optimized with one side. Based on the idea that we have above reasoned, an improved GM(1,1) model of integrated optimizing its background value and initial condition is proposed in this paper, which applies a more logical calculating formula of background value based on Newton interpolation and a relatively ingenious algorithm of initial condition by adding an arbitrary constant. Furthermore, the LCD TV annual output of China is used as a case study to examine the model reliability and accuracy. By comparisons of simulation and prediction data, a higher precision is found.