Pv layout optimization: String tiling using a multi-objective genetic algorithm

Urban environments present new challenges to the integration of photovoltaic systems onto the building envelope. This study describes a multi-objective genetic algorithm developed for the maximization of the PV electrical production and minimization of relevant system costs, such as wiring, modules and inverter costs. Two case studies are analyzed, featuring a partially shaded rooftop and a vertical facade. The PV layouts suggested by the genetic algorithm outperform the conventional configurations by reducing the cost of produced electricity between 6% and 18%. The optimization results also show that layouts with more but shorter PV strings achieve higher energy yields. However, when compared with longer clustered strings in the most sunlit areas of the surfaces, this results in the doubling of costs.