Optimized Grey Prediction Model Based on Multiple Background Projection and Sequence Decomposition

Purpose: The purpose of this paper is to enhance the forecasting performance of grey model and generate accurate prediction values of Chinese carbon emissions.Design/methodology/approach: A novel grey prediction model with multi-projection is designed. It adapts to the data characteristics, decomposes the background sequence and generates corresponding time response functions, then employs a differential evolutionary algorithm to merge the time response functions which are generated from the bi-level structure.Findings: The proposed model is tested by numeric case and applied to predict the trend of Chinese carbon emissions. The computational results show that the newly proposed multi-projection grey model performs significantly better than other three benchmark models.Originality/value: Accurate prediction of carbon emissions based on existing small sample data is of great significance for planning and decision making. This paper forecasts Chinese carbon emissions in 2020-2025 by using the latest grey prediction model. which are expected to reach 10527.51 Mt by 2025, up 6.6% from 2019, with an average annual growth rate of 1.07%. Additionally, this study proposed a framework for modeling theory that may be developed in other fields. Using a flexible operator and artificial intelligence, the degree of accuracy, acceptance, and application in and by practice might be modeled