Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

A Lightweight DL Model for Smart Grid Power Forecasting with Feature and Resolution Mismatch

Sarah Al-Shareeda, Gulcihan Ozdemir, Heung Seok Jeon, Khaleel Ahmad

TL;DR

This work tackles robust short-term smart-grid power forecasting under asymmetric, noisy inputs using a lightweight DL pipeline. It combines hourly downsampling, dual-mode imputation (mean and polynomial regression), and standardized normalization with a hybrid BiGRU-LSTM architecture to predict next-day consumption from limited test-time features. Across a real-world competition dataset, Standard scaling consistently offers the best predictive performance, yielding an average RMSE of $601.9$ W, MAE of $468.9$ W, and accuracy of $84.36\%$, with low inference latency suitable for deployment. The study demonstrates that targeted preprocessing and compact recurrent models can deliver fast, accurate forecasts even when test inputs are incomplete or mismatched, and suggests extensions to multi-step and uncertainty-aware forecasting.

Abstract

How can short-term energy consumption be accurately forecasted when sensor data is noisy, incomplete, and lacks contextual richness? This question guided our participation in the \textit{2025 Competition on Electric Energy Consumption Forecast Adopting Multi-criteria Performance Metrics}, which challenged teams to predict next-day power demand using real-world high-frequency data. We proposed a robust yet lightweight Deep Learning (DL) pipeline combining hourly downsizing, dual-mode imputation (mean and polynomial regression), and comprehensive normalization, ultimately selecting Standard Scaling for optimal balance. The lightweight GRU-LSTM sequence-to-one model achieves an average RMSE of 601.9~W, MAE of 468.9~W, and 84.36\% accuracy. Despite asymmetric inputs and imputed gaps, it generalized well, captured nonlinear demand patterns, and maintained low inference latency. Notably, spatiotemporal heatmap analysis reveals a strong alignment between temperature trends and predicted consumption, further reinforcing the model's reliability. These results demonstrate that targeted preprocessing paired with compact recurrent architectures can still enable fast, accurate, and deployment-ready energy forecasting in real-world conditions.

A Lightweight DL Model for Smart Grid Power Forecasting with Feature and Resolution Mismatch

TL;DR

This work tackles robust short-term smart-grid power forecasting under asymmetric, noisy inputs using a lightweight DL pipeline. It combines hourly downsampling, dual-mode imputation (mean and polynomial regression), and standardized normalization with a hybrid BiGRU-LSTM architecture to predict next-day consumption from limited test-time features. Across a real-world competition dataset, Standard scaling consistently offers the best predictive performance, yielding an average RMSE of W, MAE of W, and accuracy of , with low inference latency suitable for deployment. The study demonstrates that targeted preprocessing and compact recurrent models can deliver fast, accurate forecasts even when test inputs are incomplete or mismatched, and suggests extensions to multi-step and uncertainty-aware forecasting.

Abstract

How can short-term energy consumption be accurately forecasted when sensor data is noisy, incomplete, and lacks contextual richness? This question guided our participation in the \textit{2025 Competition on Electric Energy Consumption Forecast Adopting Multi-criteria Performance Metrics}, which challenged teams to predict next-day power demand using real-world high-frequency data. We proposed a robust yet lightweight Deep Learning (DL) pipeline combining hourly downsizing, dual-mode imputation (mean and polynomial regression), and comprehensive normalization, ultimately selecting Standard Scaling for optimal balance. The lightweight GRU-LSTM sequence-to-one model achieves an average RMSE of 601.9~W, MAE of 468.9~W, and 84.36\% accuracy. Despite asymmetric inputs and imputed gaps, it generalized well, captured nonlinear demand patterns, and maintained low inference latency. Notably, spatiotemporal heatmap analysis reveals a strong alignment between temperature trends and predicted consumption, further reinforcing the model's reliability. These results demonstrate that targeted preprocessing paired with compact recurrent architectures can still enable fast, accurate, and deployment-ready energy forecasting in real-world conditions.
Paper Structure (14 sections, 2 equations, 2 figures, 3 tables)

This paper contains 14 sections, 2 equations, 2 figures, 3 tables.

Table of Contents

  1. Introduction and Background
  2. Lightweight Power Forecasting Model Design
  3. Input Data Description
  4. Data Preprocessing
  5. Downsizing of Training and Validation Data
  6. Imputation of Test Data
  7. Normalization Step
  8. Hybrid GRU-LSTM Forecasting Model
  9. GRU-LSTM Architecture
  10. Model Training and Validation
  11. Simulations and Analysis
  12. Evaluation Metrics
  13. Results and Discussions
  14. Conclusion and Future Work

Figures (2)

  • Figure 1: Positioning our contribution within the forecasting landscape. Bridging it with real-world data and deployment challenges.
  • Figure 2: Block diagram of the proposed lightweight forecasting pipeline.