Abstract: Aiming at existing defects in backpressure value setting of most air-cooled units in China, by taking advantage of the parameters obtained during the operation of a 300MW direct air-cooled unit, a multi-objective optimization genetic algorithm (GA) was used to establish a specific mathematical model for back pressure and air cooler fan power consumption, and then solve the optimal solution for back pressure and minimum air cooler fan power consumption under certain constraints. Through this dynamic setting method, the backpressure setting of air-cooled units are optimized dynamically under variable load and AGC conditions, which is of great practical significance for the operating parameter adjustment of air-cooled units as well as the control strategy optimizations. It is also beneficial for the safe and economical operation of air-cooled units.