Abstract: The non-black-start generator start-up sequence optimization is an essential step for power system restoration. a multi-objective optimization model with consideration of preferences in terms of the relative importance of objectives is built to optimize generator start-up sequence and associated restoration paths simultaneously. The a priori elicitation method is used to solve the multi-objective optimization problem. The preference information is explicated by designing a subset of the whole feasible space. Then, the preferred solutions are searched in the designed region by using DE-EDA. Several Pareto optimal generator start-up sequences and restoration paths which are more desirable for the preferences can be obtained after the model is solved. Finally, a practical power system, the western power system of Shandong province in China is employed to demonstrate the effectiveness of the proposed method.