Optimal Placement of Grid-Connected Solar Photovoltaic Systems Using Artificial Intelligence Methods

• Main Authors: Ansari, Sam, Hamid, Abdul-Kadir, Hindawi, Noor Ahmad Al
• Format: Conference Proceeding
• Language: English
• Subjects: Climate change; clustering; Clustering algorithms; data mining; grid-connected solar photovoltaic systems; Image processing; machine learning; Machine learning algorithms; Photovoltaic systems; Renewable energy sources; Simulation; Temperature
• ISSN: 2161-1351
• DOI: 10.1109/DeSE54285.2021.9719570

Solar energy is a sustainable, clean, and free energy source used to supply heat, electricity, and even fuel and chemical energy to residential, commercial, and industrial centers. The problems associated with fossil resources and the consequences of environmental and global climate change have created good opportunities for solar energy to compete with fossil fuels, especially in countries with high radiation potential. Scientific and technical weaknesses, variations in irradiance level due to climate and seasonal changes, radiation angles, geographical location, etc., have confined the solar energy-related applications. The optimal placement of the photovoltaic (PV) power stations providing maximum performance is critical to be addressed. This study intends to empower grid-connected solar PV systems by investigating various constraints and influencing factors related to the location of solar farms. We exploit selected machine learning algorithms such as k-means, k-medoids, fuzzy c-means (FCM) methods, as well as a new proposed algorithm based on an image processing approach to optimize the system by detecting the appropriate clusters and site locations. Finally, by performing several simulations, we compare the results and the efficiency of the algorithms. Our results further designate the significant superiority of the proposed techniques.