Perpetual solar potential of a village by machine learning and feature extraction in UAV

Abstract

The process of sustained rural development mandates uninterrupted supply of electricity and water without dependency of the urban infrastructure. Rural areas are often neglected causing power cuts for hours in villages particularly during summer. The study area, a village very near to coast, has high potential for harvesting solar energy. Scientific investigation of the feasibility and utilization of solar irradiance is to be evaluated, to identify the potential hotspots. UAV based Remote Sensing approach associated with socio-economic characteristics, was adopted for the study. A field survey was carried out for assessing the power consumption patterns of individual households. Feature extraction was performed using machine learning technique resulting in faster and efficient extraction of spatial information from UAV. A Regression analysis is carried out to correlate household power consumption with the physical and economic characteristic variables of individual buildings. The cumulative quantity of solar irradiance was used to quantify solar energy harvesting potential using UAV to promote sustained source of green energy for the village. A suitability analysis was performed for optimal location of the community solar power plant to make the village self-sufficient with respect to Electricity. The economic feasibility of the power plant was carried out to demonstrate the feasibility of the project.