To enable renewable energy sources (RES) to better and more efficiently assist traditional generators in attending the automatic generation control (AGC), firstly, renewable energy power plants and energy storage systems (ESSs) are formed as the aggregated generation unit (AGU). Secondly, a real-time mesh adaptive multi-objective optimization algorithm (MA-MO) calculation is proposed. The proposed algorithm combines the real-time simulation and MA-MO results based on the future load disturbance prediction data. Then the optimal allocation ratio of AGC instructions is calculated for the 1-minute-ahead operation of the traditional generator and AGU, improving the AGC performance with the minimum system frequency deviation and smoothing the state of charge (SoC) of ESS. Finally, the proposed optimization algorithm is verified based on the simulation of a field AGU system.