摘要 根據預防維護之相關文獻,大部分屬於無限時間域的研究,其假設置換後的新設備與被汰換的舊設備具有相同條件的成本與特性。然而,科技日新月異,設備汰舊換新的速度快速,新舊設備在其壽命週期內具備的失效過程、效能、成本與維護效益等特性差異頗大,此無限時間域的假設較無法符合實務狀況。因此,本研究以有限時間域為研究範圍,針對維護效益為失效率縮減型,建構有限時間域的週期型預防維護模型與逐次型預防維護模型,分別考慮保固期,以及在可使用壽命終止時之轉售價值,以期望總維護成本最小化為目標函數,找出最佳預防維護政策,並提出最佳解定理與最佳解演算法,再以韋伯失效分配為例,進行敏感度分析,分別探討並比較週期型與逐次型二種預防維護模型之差異。
Abstract There are few papers in the preventive maintenances area focusing on the finite time span issue. The literature has shown that most researchers are devoted to studying the maintenance problems in the infinite time span which assume that the replaced system in each replacement cycle has the same conditions and costs. However, the useful operating life for most systems is finite in real world. Moreover, the conditions of a new equipment, such as failure process, efficiency, cost, maintenance, and etc., may be very different from the conditions of the equipment of the previous replacement cycle. It turns out that the hypothesis of infinite time span for the preventive maintenances problem does not suitable for the practical situations. Therefore, the purpose of this research is to develop an optimal periodic and sequential preventive maintenance policy with failure rate reduction in a finite time span under a given warranty period and a re-sale value by minimizing the expected total maintenance cost. In this research, the algorithms and theorems for searching the optimal solutions are presented. Examples with Weibull failure cases are given and the sensitivity analysis of optimal solution is also compared and provided.