Short-Term Electrical Load Forecasting Through Optimally Configured Long Short-Term Memory
Short-term electrical load forecasting plays a pivotal role in modern energy systems, addressing the need for accurate predictions of electricity demand within a time frame ranging from a few hours to a few days. Inaccurate predictions can lead not only to operational challenges but also to economi...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
OICC Press
2024-12-01
|
| Series: | Majlesi Journal of Electrical Engineering |
| Subjects: | |
| Online Access: | https://oiccpress.com/mjee/article/view/7992 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Short-term electrical load forecasting plays a pivotal role in modern energy systems, addressing the need for accurate predictions of electricity demand within a time frame ranging from a few hours to a few days. Inaccurate predictions can lead not only to operational challenges but also to economic and environmental consequences, highlighting the critical importance of short-term electrical load forecasting in today’s energy landscape. This research aims to mitigate these issues by developing an optimally configured Long Short-Term Memory (LSTM) model for short-term electrical load forecasting in Tamil Nadu, specifically targeting the Villupuram region in India. Although LSTM models are known for their effectiveness, achieving optimal performance in short-term load forecasting requires a tailored approach. Hyperparameter optimization is essential for configuring the LSTM model for this purpose, as manual or trial-and-error hyperparameter tuning is time-consuming and computationally intensive. To address this challenge, this research
integrates the Cauchy-distributed Harris Hawks Optimization (Cd-HHO) method to optimally configure the LSTM model. The Cd-HHO-optimized LSTM consistently achieves lower Mean Squared Error (MSE) than other state-of-the-art methods, with MSE values of 0.7225 in the 2017 dataset, 0.974 in the 2018 dataset, and 0.116 in the 2019 dataset.
|
|---|---|
| ISSN: | 2345-377X 2345-3796 |