Applications of Fractional Time Delayed Grey Model in Primary Energy Consumption Prediction

Abstract

The prediction of total energy consumption is crucial across various domains including the economy, environment, market, and geopolitics. Accurate forecasts can guide policy-making, investment decisions, and international strategies, contributing to sustainable development and energy security. Fractional models have been proven to better capture the long-term memory effects and complex dynamic characteristics of systems, with time delay playing a crucial role in capturing dynamic behaviors. Such models enhance the accuracy and reliability of predicting future trends and behaviors. For the prediction of primary energy consumption in South and Central America, the Middle East, and Africa, this study opts for the existing fractional time delayed grey model, optimizing the fractional order using the particle swarm optimization algorithm. Experimental results demonstrate that in most cases, the predictive capability of the fractional time delayed grey model surpasses that of other grey models. This indicates the effectiveness and reliability of the model in forecasting energy consumption, providing valuable references and foundations for decision-making in relevant fields.