In the context of increased development of renewable energy,more and more researches are seeking to tap the potential of new energy development so as to provide data support for long-term energy planning studies,scheduling simulations and energy policies,while the complexity of urban buildings increases the difficulty in assessing photovoltaic potentiality.In this paper,the existing research methods of photovoltaic potential are introduced in detail.The deep learning-based semantic segmentation model is used to extract building roofing according to high-resolution satellite images.And then,the photovoltaic potential of Shaoyang City in Hunan Province is systematically evaluated.Finally,the potential of distributed PV power in some areas of the 14th Five-Year PV Plan is estimated.
distributed photovoltaic powermachine learningurban renewable energy planningsatellite image
国家科技期刊平台
NSTL
万方数据
