Prediction of the lahore electricity consumption using seasonal discrete grey polynomial model.

The article focuses on forecasting seasonal fluctuations in electricity consumption in Lahore, Pakistan, using a novel Seasonal Discrete Grey Polynomial Model, denoted as SDGPM(1,1,N). Based on quarterly electricity consumption data from 2014 to 2021 provided by Lahore Electric Supply Company (LESCO), the SDGPM(1,1,N) model incorporates seasonal adjustment factors to improve prediction accuracy. Comparative analysis with nine other models—including the original discrete grey polynomial model DGPM(1,1,N), statistical models (LR, AR, ARIMA), deep learning models (LSTM, RNN, GRU), and exponential smoothing models (SES, Holt-Winter)—demonstrated that SDGPM(1,1,N) significantly outperforms these alternatives in capturing seasonal patterns and forecasting accuracy. The model predicts a gradual increase in Lahore’s electricity consumption from 2022 to 2025, with continued significant seasonal variation, providing valuable insights for policymakers to better balance supply and demand amid Pakistan’s ongoing electricity shortages.