Pv Potential Analysis Through Deep Learning and Remote Sensing-Based Urban Land Classification

Abstract

Considering land use types is crucial when evaluating photovoltaic (PV) potential in a city as it enables optimal space allocation, proximity to energy demand centers, and improves energy efficiency. However, previous evaluations have overlooked urban land use types. To address this gap, this study proposes a framework utilizing remote sensing data and deep learning-based methods to achieve eight fine-granularity and three coarse-granularity land use classifications. The framework calculates the PV installation area for each land use type and evaluates their power generation potential. Case studies in two cities demonstrate that Heilbronn land is suitable for ground PV installations, and rooftop PV installations are the most productive for electricity generation in Christchurch. Finally, the uncertainty of the PV installation ratio by adopting σi and the confidence interval of potential estimation is discussed. This work experiments with the framework successfully and highlights the effects of the PV installation ratio on the power generation of each land use, providing valuable instructions for urban land utilization and PV installation.

Keywords: Solar irradiance, PV potential, Remote sensing, Land use, Classification