The steam power system in steel industries is featured by a variety of energy and various energy producers. In this paper

a method to convert energy into electricity equivalent is introduced to analyze the efficiency of energy conversion in the system

and mathematical programming and optimization are employed to improve energy management based on the background of steam power system in a large iron and steel company. With the establishment of mixed integer nonlinear programming model which is multi-objective

the global optimal solution will be achieved by solve the multi-objective-constrained model by LINGO step by step on condition that the objectives

the minimum cost of energy hourly and the maximum exergy efficiency

and the constraints including the capacity of power plants

energy demand and operating cost are set. Pareto front is to be achieved by solving the objective function-the maximum exergy efficiency-with cost constraint by the minimum cost of energy hourly multiplying an over-relaxation which is little bigger than 1. Rationality

feasibility and efficiency of ultimate optimization solution by stepwise multi-objective optimization is to be demonstrated in comparison with solutions by single-objective optimization and multi-objective genetic algorithm. The schedule of effective operation at low cost of steam power system is to be well-founded in theory in accordance with the optimal solution.