Shape detection algorithm applicable to solar estimation

This paper presents a bio-inspired hybrid algorithm for shape detection applicable to solar estimation in solar power plants. The objective is to locate and characterise the shape of a cloud over a solar power plant based on low level irradiance measurement with a small fleet of Unmanned Aerial Vehicles (UAVs) equipped with direct normal irradiance sensors. Toe hybrid algorithm takes inspiration and adapts ideas of the ant colony optimisation algorithm (ACO) and also uses a standard cover area algorithm, separating the field into two grids, one for each layer of the algorithm, to find the area affected by the cloud. Once the low irradiance zone is located by one of the UAVs, the others go to help it. This team delimits the cloud border using concepts of an image processing technique. Finally, the algorithm is tested by simulations.