Abstract: The recursion relationship of failure rate between preventive maintenance cycles is built up, and the non-linear optimal model of equipment preventive maintenance cycle in finite time horizon is proposed considering repair cost, preventive maintenance cost and production loss cost. This approach avoids the shortcoming of static model in infinite time horizon. As an analysis example of the model, the equipment failure distribution following Weibull distribution is optimized by Genetic Algorithm (GA) approach. The result shows GA is an effective approach to obtain the global optimal cycle in limited time. This model can be used to provide information and decision supporting for the maintenance plan programming and job shop scheduling.

Key words: preventive maintenance, age reduction factor, finite time horizon, failure rate

摘要： 建立了预防性维修周期间故障率的递推关系式，给出了有限时间区间的设备预防性维修策略的非线性优化模型。该模型综合考虑了维修成本、预防性维修成本和生产损失成本，克服了无限时间区间稳态分析操作性差的缺点，并以故障分布形式为威布尔分布的设备为例，用遗传算法进行优化。计算结果显示，遗传算法能以极快的收敛速度达到全局最优，具有较高的计算效率。模型可为维修计划的制定和现场的作业调度提供决策支持和信息支持。

关键词: 预防性维修, 役龄回退因子, 有限时间区间, 故障率