Enhancing Short-Term Load Forecasting Based on a Hybrid Model of Prophet and Temporal Fusion Transformer

Efficient short-term load forecasting plays a crucial role in ensuring the stability and reliability of power systems, enabling utilities to make informed decisions regarding generation, transmission, and distribution. In this paper, we emphasize the significance of accurate short-term load forecasting highlight the importance of accurate short-term load forecasting, and highlight the challenges associated with traditional forecasting methods. We have discussed the importance of two prominent forecasting models, Prophet and Temporal Fusion Transformer (TFT), in addressing these challenges. Prophet, known for its robustness in handling time series data and capturing seasonal patterns, offers valuable insights into trend and seasonal components. TFT, a deep learning-based model, excels in capturing complex temporal dependencies and dynamic patterns in the data. By integrating these models, we aim to harness the strengths of both approaches, leveraging Prophet’s interpretability and TFT’s predictive power. We discuss the potential benefits of this combination, including improved forecasting accuracy, enhanced adaptability to varying data patterns, and the ability to capture both short-term fluctuations and long-term trends in load demand, from the results we have discussed that the hybrid approach of combining the Prophet and TFT model gives more accurate predictions compared to other models.