This paper proposes a strategy base on a Population-Based Incremental Learning (PBIL) algorithm to reduce the fuel consumption in hybrid fuel cell/photovoltaic power systems. The strategy is focused on increasing the power produced by the photovoltaic (PV) generator to reduce the fuel needed to supply the load. Such an improvement is achieved by reducing the effect of shadows over the system, which is done by dynamically reconfiguring the connections between the PV modules. In such a way, the proposed PBIL algorithm accelerates the detection of the optimal configuration, in comparison with classical approaches. Finally, detailed simulations are used to demonstrate the efficiency of the proposed solution.