Establishment of a grey inhomogeneous gas demand forecasting model

To predict the future gas demand accurately, the changes of natural gas sales volume data over China in recent years are numerically analyzed. Deficiencies exist in the predictions of traditional gray forecasting model. Therefore, to achieve the consistency of sequence data type that can be predicted with the actual sequence data type, through optimizing the model structure, a gray inhomogeneous model which considers the inhomogeneous exponential law change of original sequence is established, and its feasibility is proved using least squares method and matrix operations. Combing the white solution and exact solution of the model, the underlying causes and influencing factors for instable forecasting results are discussed in detail. Taking the Shaanxi Province as an example, the model is validated and used for forecasting. The results show that the gray inhomogeneous model has small error and highly fit original data type. The model is quite applicable to original sequence with inhomogeneous exponential law change.