An oil imports dependence forecasting system based on fuzzy time series and multi-objective optimization algorithm: Case for China

Oil production and consumption is of great importance for the sustainable development and management of energy and environment. The forecasting of oil imports dependence caused by the gap between production and consumption is particularly crucial in the strategic deployment of oil development. However, researches on oil import dependence forecasting are often limited by the small size of data samples and assumptions, and the previous single-objective optimization algorithms only focus on the improvement of forecasting accuracy but ignore the stability. Therefore, in order to overcome the shortcomings of researches, in this paper, a hybrid forecasting system for oil imports dependence forecasting based on fuzzy time series and multi-objective optimization algorithm is proposed considering the accuracy and stability simultaneously to achieve the balance and optimality and bridge the limitations of small sample forecasting. The proposed forecasting system is compared with other small sample forecasting models and the fuzzy times series model with traditional interval partition methods. The results show that the proposed system is superior to the traditional methods in all indicators for oil import dependence forecasting. Otherwise, the out of sample forecasting results provided in our research indicate that the oil import dependence will maintain an upward trend, but the rate of increase will slow down. The research results can not only provide the basis for the planning and control of oil import and safety, but also benefit the perceptions of oil price trend in the world energy market and the adjustment of energy market structure.