New tool to estimate cities’ rooftop PV potential considers roof superstructures
- Admin
- Aug 2, 2024
- 1 min read
A group of scientists from Germany have created a new deep learning-based method for city-scale PV rooftop potential which considers roof superstructures to alleviate overestimation. The system named SolarNet+ uses CNN to learn roof orientation and superstructure maps, achieving high prediction accuracy. The framework was tested in Wartenberg and Munich, showing impressive results with room for improvement. Future research includes improving transferability by collecting more training samples from various cities and implementing domain adaptation techniques. Finally, the system was integrated with various LCZ types and tested in Brussels, revealing high rooftop solar potential efficiency in specific urban types.
Source:
Want to read more?
Subscribe to vitaliydaviy.wixsite.com to keep reading this exclusive post.
