Optimization of PV Battery Systems Using Genetic Algorithms

AbstractA modular simulation model of a PV battery system has been developed and integrated into a genetic algorithm framework in order to evaluate optimal sizing of such systems under various boundary conditions. The presented paper describes the simulation assumptions and presents optimization results for a PV battery system having a DC topology, comparing current economic scenarios with and without KfW funding. Sensitivity analyses provide information on critical boundaries to reach economic operation. The fitness of a system is evaluated based on the levelized cost of electricity (LCOE) defining the average cost – including all investment and operation cost over the system lifetime – per kWh supplied to the load