Abstract: In this paper, a new fractional-order grey forecasting model with a temporal power term that can handle both annual and quarterly forecasting tasks for GDP is presented. The model's characteristic is that it has a dynamic simulation parameter, which can automatically adjust the structure of the model according to the need of the prediction task to achieve the purpose of accurate prediction. In addition, the fractional order parameter and power term parameter of the model play an important role in enhancing the adaptive performance of the model. In particular, an excellent intelligent optimization algorithm, the Ant lion optimizer, is used to solve the model's programming model to obtain the hyperparameters for modeling quickly. In this study, China's annual GDP and quarterly GDP are used as research objects to verify the validity of the new model. The experimental results show that all evaluation indicators of the proposed method are better than those of its competitors. Therefore, the model has some application value. 

Key words: Fractional grey forecasting model , GDP forecasting , Ant lion optimizer