System-on-chip implementation of a PV dynamical reconfiguration algorithm

Dynamical reconfiguration is receiving attention from the scientific community because of the advantages it can ensure with respect to module-dedicated electronics when partial shadowing of the photovoltaic field occurs. Indeed, lower cost and conversion losses as well as improved diagnostic capabilities make this solution of high interest for industrialization in a near future. Some stochastic algorithms have been proposed recently for determining the best electrical connection among the panels that ensures the highest power production for the actual shadowing pattern. In this paper, no novel reconfiguration algorithm is proposed, but the optimized implementation of one of such recently proposed approaches is described. The advantages achieved by using a system-on-chip platform, through a suitable exploitation of the features it offers, are highlighted. The experimental results confirm that the use of a system-on-chip platform gives significant benefits for the proposed application. Indeed the parallelization of the algorithm allows to reduce the execution time greatly with respect to what can be achieved through a standard software implementation.