With the rapid development of renewable energy, traditional power generation such as combined cycle unit needs to startup or change load quickly to balance the variability brought by renewable resources. In the combined cycle unit, the startup of steam turbine has a great influence on the whole set, and it is necessary to ensure safety, shorten the startup time, and reduce pollutant emissions. These optimization objectives are inter-related and have complex contradiction, which is the difficulty for multi-objective optimization. Costs function consisting of startup costs, fatigue life damage costs, pollutant emissions costs, and revenue gained from electricity is proposed to balance the contradiction more objectively. Finite exhaustive method (FEM) combining thermal-structural finite element calculation and multi-objective optimization is proposed for multi-objective startup process optimization, and its results are compared with those of minimum startup time and minimum fatigue life damage optimization. It can automatically search for a unique global optimal solution for engineering practice, rather than solution sets obtained by Pareto Optimality, which is beneficial for application in different combined cycle steam turbines and startup process. Multi-objective optimization scheduling shortens startup time from 105 to 93 min, reduces maximum stress from 493 MPa to 440 MPa, and reduces costs function by 66.6%. The comparison with the multi-objective optimization results of the response surface method (RSM) proves the reliability and validity of this method. The practical inspection results prove that the optimal scheduling is safe and effective.