Abstract:

In view of the existing problems in forecasting for small sample oscillatory sequences, this paper puts forward a modeling approach for oscillatory sequences based on the first-order single-variable grey power model (abbreviated as GM(1,1) power model). A non-linear optimization model is constructed based on the relationships between parameters in order to get the best value of the power exponent and the highest forecasting precision. The results show that the proposed modeling approach can reflect the fluctuation of the raw data and is easy to solve using a computer. The range of application of the grey model is further extended. Finally, the validity and practicability are illustrated with two examples.