In this study, virtual inertia control-based vehicle-to-grid (V2G) concept for the secondary load frequency control of AC islanded microgrids (MGs) is presented. Since autonomous modern power grids typically have much smaller inertia in comparison to conventional grids, a very fast and high-performance control plant is necessitated for controlling reserved power plants (e.g., batteries) to overcome any voltage and frequency instability. This issue can be solved by the short-term battery energy storage system, inverter, and proper inertia control technique. Because of the high cost of the battery energy storage system, utilization of controllable loads like electric vehicle (EV) to low-inertia modern power grids is presented for the decreasing of the needed capacity of the energy storage system. Hence, an optimal non-integer model predictive control is applied for the secondary frequency regulation in low-inertia stand-alone MGs with the V2G technology. Since the performance and accuracy of the applied non-integer controller are depended on its parameters’ value, a modified heuristic optimization algorithm is proposed for tuning the controller's parameters. In the end, the validity and feasibility of the proposed method are investigated by the model-in-the-loop experimental results.