Owing to the development of renewable energy sources and reduction in diesel consumption, the power supply cost in outlying islands can be minimized by installing solar photovoltaic (PV) systems. However, the island power grid usually has lower inertia, limiting the PV hosting capacity. Integrating a virtual synchronous generator (VSG) control with an energy storage system (ESS) and PV is beneficial to strengthening inertia. Nevertheless, incomplete control parameters may trigger oscillations when multiple power sources are connected. This study proposes a parameter selection method for ESS-VSG and PV-VSG that adopts a multiobjective genetic algorithm (MOGA) optimization method to simultaneously increase the frequency nadir and minimize the settling time after a disturbance. Using the controller design, the energy dispatch can comply with the PV power-priority increase principle when the frequency is regulated-up, and ESS priority charge when the frequency is regulated-down, to improve energy efficiency. In addition, the actual data of an outlying island located in the Penghu Archipelago in Taiwan is modeled via the DIgSILENT PowerFactory to interact with MOGA. Furthermore, the economic impact of risk factors due to power curtailment is analyzed. As a result, the frequency nadir can be increased by 0.6 Hz under disturbance and the annual power supply cost of the whole island is reduced by 6%.