AN INNOVATIVE GREY MODEL APPROACH WITH FRACTAL CALCULUS FOR FORECASTING RENEWABLE AND WASTE-BASED ELECTRICITY GENERATION

This study proposed and employed a Fractal Grey model to estimate the fraction of electricity produced from renewable and waste sources in Türkiye from 2010 to 2024. It was compared to the classical grey model and several nonlinear grey models and a sensitivity analysis was performed with different training–testing periods. The results indicate that the proposed model can effectively represent the small changes in the short and limited dataset and has a higher forecasting accuracy and more stable parameters than the other models. The traditional grey model and Discrete Grey model are both good at forecasting the trend, but they are slower to adjust to sudden changes than the Fractal Grey model. The sensitivity analysis also shows that the parameters of the Fractal Grey model are not sensitive to the changes of the training period, and the model is relatively robust to the changes of the range of data. Electricity generation projections for 2025–2030 generated by the model indicate that electricity generation by renewable and waste energy sources will share a bigger percentage of the total electricity generation in Türkiye. The results indicate that the Fractal Grey model is useful for energy planning and policy-making and can be a reliable model when data are limited and there is some uncertainty. In summary, the study offers a scientific reference for the energy transition strategies, capacity planning, and investment decisions that are taken in Türkiye, and also establishes the Fractal Grey model as an innovative approach that can easily provide accurate forecasts even with a short dataset.