Abstract:  The traditional periodic maintenance strategy often led to under-maintenance and over-maintenance of the ship equipment. To improve the scientificity of maintenance decision， a ship grade repair decision method was proposed based on the concept of preventive maintenance. According to the technical status and maintenance information of ship equipment， the failure mode and effectiveness analysis of key important components were carried out to determine their repair demand level and repair scope； meanwhile， the Weibull distribution model and the maximum likelihood estimation method were used to predict the median life of the equipment， and the production rules were defined； then， based on the three-level repair BOM structure of ship-system-critical parts， logic decision and rule reasoning were conducted from bottom to top to determine the repair grade and timing of the entire ship. A ship grade repair decision system was developed and functionally verified by cases， which might meet the application requirements of autonomous decision for ship grade repair and provide a feasible way for preventive maintenance of ships. 

Key words: preventive maintenance, grade repair, maintenance decision, failure prediction, Weibull distribution, ship equipment

摘要： 传统的定期维修策略常导致舰船装备欠维修和过维修。为提高维修决策的科学性，提出了一种基于预防性维修理念的舰船等级修理决策方法。根据舰船装备的技术状态和维修信息，对关键重要部件（简称关重件）进行失效模式和效果分析，确定其修理需求级别和修理范围；同时采用威布尔分布模型和极大似然估计法预测装备的中位寿命，并定义产生式规则；进而基于舰船-系统-关重件三级维修BOM结构，自下而上地进行逻辑决断和规则推理，确定整舰的修理等级和时机。在此基础上开发了舰船等级修理决策系统，并通过案例进行功能验证，该系统可以满足舰船等级修理自主决策的应用需求，为舰船的预防性维修提供了一种可行的途径。

关键词: 预防性维修, 等级修理, 维修决策, 故障预测, 威布尔分布, 舰船装备