In system health management, the extreme value of reliability, availability or maintenance cost is usually used as a criterion for determining maintenance intervals. However, with the complexity of the equipment system and the diversification of decision makers’ needs, the single objective maintenance decision-making problem can no longer satisfy the current situation. Therefore, a multi-objective maintenance decision model based on the delay time theory and considering imperfect maintenance is established for the periodic detection strategy, and the multi-attribute utility theory is used to solve the model. Firstly, the reliability, availability and maintenance cost models are established then the utility functions of each attribute are constructed. Then the weighting method based on technique for order preference by similarity to ideal solution thought is used to assign the weight of the attribute, and the comprehensive utility value of different maintenance period is obtained by weighting, and the maintenance interval is determined according to the utility value. Finally, the validity and feasibility of the proposed method are verified by an example analysis.