Optimizing renewable energy site selection in rural Australia: Clustering algorithms and energy potential analysis

Renewable energy development is a critical issue in Australia, and identifying suitable regions for constructing renewable energy plants is an essential step towards achieving sustainable energy goals. This work presents insights and techniques aimed at identifying optimal locations for renewable energy stations in rural areas across Australia as a whole. Following the above-mentioned idea, the study uses clustering algorithms to explore the optimization of renewable energy site selection. The research focuses on applying these algorithms to analyze spatial data and identify optimal geographic clusters for potential development based on technical parameters like solar irradiance and wind speed. Various clustering algorithms were employed in line with our methodology, namely K-Means, Density-Based Spatial Clustering of Applications with Noise (DBSCAN), Hierarchical clustering, and K-Medoids. Each algorithm generated clusters, facilitating the identification of appropriate regions based on a range of data attributes. A genetic algorithm was integrated into an iterative process to identify the most appropriate clustering method. Additionally, The HOMER Pro software was used to process the generated cluster centers and estimate the solar and wind energy potential for each location. The analysis revealed that solar panels consistently outperform wind turbines in energy generation across various clusters and algorithms. While the genetic K-Means algorithm performed best based on clustering evaluation metrics, the genetic K-Medoids algorithm produced the highest energy output. However, the latter incurred the highest financial costs, highlighting a trade-off between energy production and economic feasibility. This study provides valuable insights into the application of clustering techniques for renewable energy site selection and identifies challenges and limitations that require further investigation.