For the large-scale complex system

the photovoltaic output fluctuates greatly

and the state of charge (SOC) and the depth of discharge are inconsistent. A control strategy of optical storage power station based on multi-agent and particle swarm optimization is proposed. The output power objective function of the optical storage combined power station is established. The particle velocity in the algorithm is expressed by the charge and discharge power of each PCS in each energy storage unit area

and the particle position is represented by the SOC corresponding to each PCS. Through continuous optimization of constraints and objective functions

the optimal solution is found. The simulation results of Matlab/Simulink show that the proposed multi-agent particle swarm optimization algorithm effectively reduces the power fluctuation rate and optimizes the SOC of battery corresponding to the main agent of the optical storage power station.