L R AS Published on Saturday 5 September 2020 - n° 330 - Categories:new development
For cleaning low-power photovoltaic power plants
Scientists from the Sorbonne, the École Normale Supérieure de Rennes (ENS Rennes) and the University of Paris-Saclay have developed a machine learning model for cleaning low-power photovoltaic power plants and stand-alone solar panels in rural areas isolated from the grid.
The system is based on a
620 W, a motor pump equipped with an inverter with Maximum Power Point Tracking (MPPT) and an 11.4 m3 water tank. Cleaning in Africa is usually done manually. It is more frequent during the dry season from November to May than during the rainy season from April to October. Manual wet cleaning is used because it is a good compromise between expensive water jets and inefficient manual dry cleaning.
In some areas, the efficiency of solar panels can deteriorate by as much as 6% in a day and up to 18% in a month if dust is not regularly removed from the panels.
The aim is to offer companies solutions that enable them to improve the maintenance of photovoltaic installations in rural areas. To achieve this, the device must become even more robust and reliable.
https://www.pv-magazine.com/2020/09/03/machine-learning-for-pv-module-cleaning/
PV Magazine of 3 September
