Compressed air energy storage system can effectively reduce the wind abandonment phenomenon caused by the randomness of wind energy, but its dynamic response time is long, and the unreasonable configuration of storage scale will affect its development. For this reason, firstly, historical data are used to obtain the typical hourly power distribution of wind power generation; then, considering factors such as user load demand, system investment cost, system maintenance cost and power sales revenue and environmental protection benefit, a model with CAES system charging and discharging power and storage capacity as constraints and maximum benefit as goal is constructed, and the improved whale algorithm is used for solving. Finally, the effects of rated capacity and rated power of compressed air energy storage system on the maximum return of the system are analyzed to verify the reliability of the algorithm. The simulation results show that for a factory customer with a typical hourly load power demand of 3.240 MW, it is economically optimal for the wind farm to keep four wind turbines running daily and configure a CAES system with a rated power of 1 MW and a rated capacity of 6.5 MWh, which reduces the amount of wind abandonment by 5.55 MWh, saves the cost of purchased power by $4,332.3, and realizes the maximum daily net benefit of $756.86.