Simulations of solar optimizers in parallel coupling

Abstract

The main objectives of this paper is to verify, for the first time, the possibility to model a solar optimizer developed in laboratory and connected to an un-shaded photovoltaic (PV) module, where his behaviors have been simulated by a commercial software, the Proteus Design Suite of Labcenter Electronics. For two different PV modules (a roof PV module and a PV module dedicated to 12V battery recharge) software simulations of I-V characteristics are compared to experimental curves. The very small differences between simulation and experimental results lead us to consider that such modelling is realistic for simple configuration, namely a solar optimizer supplying a single load. In a second time an association of two solar optimizers by output parallel coupling is designed and simulated with Proteus software. Large data files produced by Proteus software simulations require to reduce duration of studied phenomena, or else errors like insufficient memory may appear. Simulated behaviors as current injection of small PV module may be encountered when the physical conditions are not gathered, which is very astonishing, while other behaviors comply perfectly with physical reality. So, to verify the real behavior of parallel coupling an experiment is designed to highlight variations of electrical variables like PV module voltage and current without forgetting load voltage. This last experiment confirms effectively how PV modules with different specifications influence each other. Partial mismatch between simulation and experiment implies future strengthening in modeling process with Proteus software, expecting to verify that reason is not the complexity of the parallel association when electronical circuit begins to be heavy to treat. As other tracks of investigation, DC-DC converter model has to be improved, particularly with a more precise characterization of the autotransformer which is the key component of high voltage step-up. Finally, additional experiments have to be also planned with various loads (pure resistors and electronical load) to confirm or deny phenomena encountered when using particular load as incandescent lamps.