Bhardwaj et al. Res. Trends Int. J. Technol. Innov., July - September 2026, 1 (3) : 58-65
1Department of Electrical Engineering, Delhi Technological University, New Delhi, India; 2Department of Computer Science, NMIMS University, Mumbai, India
Article History
Accepted : 05 Jul 2026
Published : 10 Jul 2026
Publication Issue
Volume 1, Issue 3
July - September 2026
Page Number58–65
Short-term electricity demand forecasting must capture both fine-grained temporal dynamics and calendar-driven seasonality such as holidays, and is increasingly central to planning renewable generation and storage dispatch. This paper proposes a hybrid model that combines an LSTM network for short-horizon demand patterns with Facebook Prophet decomposition of trend and holiday effects, evaluated on three years of hourly demand data from a regional utility with growing renewable penetration. The hybrid model achieved a mean absolute percentage error of 2.9 percent for 24-hour-ahead forecasts, outperforming standalone LSTM (3.8 percent) and standalone Prophet (4.6 percent) baselines.
Keywords - load forecasting, LSTM, Prophet, renewable grid integration, energy demand
Grid operators integrating growing shares of renewable generation require accurate day-ahead demand forecasts for unit commitment, reserve planning and renewable curtailment decisions, and forecasting error directly translates into either costly reserve over-procurement or reliability risk, motivating continued refinement of hybrid statistical-and-deep-learning approaches.
Three years of hourly demand data from a regional distribution utility with growing renewable generation share was used to train a hybrid forecasting pipeline in which Prophet first decomposed the series into trend, weekly and holiday-effect components, with the residual series then modelled by a two-layer LSTM network, and forecasts reconstructed by summing the Prophet trend-seasonal component with the LSTM residual prediction.
The hybrid LSTM-Prophet model achieved a MAPE of 2.9 percent for 24-hour-ahead forecasts on the held-out test year, compared with 3.8 percent for a standalone LSTM and 4.6 percent for standalone Prophet, with the largest relative improvement observed around public holidays where Prophet calendar effects corrected LSTM under-forecasting.
Decomposing calendar-driven seasonality explicitly before applying sequence modelling improves short-term electricity demand forecast accuracy over either method alone, supporting more reliable renewable grid integration planning. Future work will incorporate temperature forecast uncertainty into the pipeline.
[1] Taylor S. J. and Letham B., Forecasting at scale (Prophet), The American Statistician, 2018. [2] Hong T. and Fan S., Probabilistic electric load forecasting: A tutorial review, International Journal of Forecasting, 2016.
© 2026 The Author(s). Published by IJEIA Editorial Office. This is an open access article under the Creative Commons Attribution 4.0 International License (CC BY 4.0).
Aakash Bhardwaj, Neelam Choudhary (2026). Hybrid LSTM-Prophet Forecasting of Electricity Demand for Renewable Grid Integration Planning. IJEIA, 1(3), 58-65.