Advancing Solar Energy Potential Assessment of Urban Landscapes: a Deep Learning and Computer Vision Based Architecture for High-Resolution Topographic Data Incorporating Temporal Shadowing
Abstract
Accurately assessing solar energy potential is essential for effective urban planning and successful solar panel installations. While conventional methods consider factors like climate conditions, low-resolution topography, and mean solar radiation, there is a pressing need to incorporate high-resolution topographic data and account for the temporal shadowing effect caused by neighbouring structures and trees. This research aims to address these challenges and develop an algorithmic architecture specifically tailored for densely populated tropical areas like New Delhi in India. By introducing computer vision techniques, this study pioneers a novel approach to analyzing temporal shadowing effects in urban environments. This integration significantly enhances the accuracy and efficiency of solar energy potential assessment while improving the methodology's scalability and generalizability. This study fills a critical gap by considering the temporal shadowing effect and utilizing high-resolution topographic data. A key focus of this research is data-driven decision-making in renewable energy planning. This comprehensive approach enables more informed decisions in urban planning, paving the way for sustainable and resilient cities. The findings hold great promise for transforming urban energy planning and fostering the widespread adoption of solar energy systems.
Keywords
Solar energy potential, Urban planning, Temporal shadowing effect, Rooftop pv mounting, Pv installation, Renewable energy planning, Satellite data, Gis, Computer vision, Deep learning, Solar radiation, Google earth engine, Building footprint.