Influence of TimeGAN-Generated Scenarios on Optimal Design for Renewable Energy Systems

This study explores the influence of variability of natural resources and electrical consumption on the optimal biobjective design of isolated renewable energy microgrids. The variability is generated by synthetic time-series data produced by TimeGAN. The two primary objectives are to minimize the Levelized Cost of Energy (LCOE) and the Loss of Power Supply in Hours (LPSH). The approach integrates rule-based energy management and stochastic scenario variations of photovoltaic (PV), wind turbine (WT) production, and consumption profiles. The results demonstrate how variations in input scenarios impact the Pareto-optimal front, offering insights into system design robustness under uncertainty. The findings underscore the importance of addressing variability in renewable energy systems and pave the way for future work on stochastic optimization and multi-criteria decision-making for more resilient and sustainable energy solutions.