Predicting hydropower generation: A comparative analysis of Machine learning models and optimization algorithms for enhanced forecasting accuracy and operational efficiency
Exponential global population growth and rapid technological advancements have increased electricity demand, strained the fossil fuel-reliant energy infrastructure, and exacerbated environmental issues like greenhouse gas emissions and climate change. Transitioning to sustainable energy sources is e...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-03-01
|
| Series: | Ain Shams Engineering Journal |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2090447925000401 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Exponential global population growth and rapid technological advancements have increased electricity demand, strained the fossil fuel-reliant energy infrastructure, and exacerbated environmental issues like greenhouse gas emissions and climate change. Transitioning to sustainable energy sources is essential for balancing energy needs with environmental conservation. Hydropower is a significant renewable resource due to its cost-effectiveness, low environmental impact, and capability to meet peak electricity demands. Optimizing hydropower generation is crucial for addressing economic and environmental concerns, though it requires comprehensive monitoring and understanding of energy conversion processes. Machine Learning techniques such as integrated Gradient Boosting and Categorical Gradient Boosting, optimized with Hunger Games search, Chaos game optimization, and Archimedes Optimization Algorithm algorithms, are used to forecast and optimize hydropower generation. The dataset involved hydropower generation data from 1819 records gathered from a particular watershed. The framework is designed to tackle challenges in the prediction of hydropower generation by effectively managing complex, multivariate data. Between the tested models, the integrated CatBoost- hunger Games search model brings out exceptional predictive accuracy, with a Coefficient of Determination of 0.9075 and a Root Mean Square Error of 45.2130 during testing. This examination’s contributions comprise the creation of a scalable, data-driven method for hydropower optimization, the illustration of its capability to decrease prediction errors remarkably, and its practical application in renewable energy management. These reports highlight the potential of the proposed configuration to elevate hydropower prediction accuracy and support the transition to sustainable energy frameworks. |
|---|---|
| ISSN: | 2090-4479 |