In the iron and steel enterprises, the consumption of gas, steam, electricity and other multi-media energy is the main component of the total energy consumption of the enterprise. Multi-media energy sources such as gas, steam, and electricity are strongly coupled in the steel production process. Therefore, combining the characteristics of specific boilers, steam turbines and other equipment and the operating status of the system under different operating conditions, a multi-media coupling optimization model is established. Aiming at the characteristics of the model with many constraints and small feasible region, a differential evolution algorithm with a hybrid penalty repair constraint processing method is proposed. After a comparative analysis of actual cases, combined with different working conditions, the effectiveness of the model and algorithm is verified, and the operating costs and energy consumption of steel enterprises are reduced.