Due to deterioration, system produces defective products and incurs the increasing of the operating cost. Many studies deal with how to choose proper maintenance actions for the system to minimize the objective function such as the total operating cost, the total expected discounted cost over planning horizon and etc.. However, maintenances are better executed in a preventive manner to ensure proper system operation and avoid the negative influence which the system deterioration brings.
The deterioration system considered herein is equipped with an on-line inspection instrument and computer. Such system is under periodic inspection which can identify the system state immediately. A PM policy is proposed for such a multi-state deterioration system to determine at each proper decision time what action should be taken for the next PM and when to take it. Such PM policy holds under the assumptions that:
1) The discrete time multi-state Markov chain is applied to describe the transitions of system states where the transition probabilities are assumed to change only after the completion of each PM.
2) The transition probability matrix is in upper triangular to reflect the system deterioration when under no maintenance.
3) The candidate actions except replacement are imperfect.
4) The considered operating cost between two inspection times is state-dependent only.
The differences between our policy and others are that the optimal actions and the optimal threshold state can be determined at each planned decision time based on the current system state in terms of maintenance data only. Hence, the enterprise can build up a useful PM policy by applying our proposed policy in terms of the maintenance data collected by the inspection instrument.
In order to illustrate the proposed PM policy and evaluate how much benefits such PM policy can obtain, we will choose the system whose deterioration obeys a non-homogeneous Gamma process for application.

如何降低系統因衰退所帶來的負面影響 (包含操作成本之增加與系統妥善率之下降等)，一直是製造業的重要課題。許多文獻也討論如何針對衰退系統在最小目標函數(例如期望總操作成本，維護週期內之平均成本等) 下找出最佳維護策略。較佳的維護策略應是以預防性的角度來定訂，即非在系統出現損失時才採取維護動作。而目前的科技已能做到：結合線上檢測儀器與電腦，對系統進行即時監控且可以快速確認系統狀態並蒐集維護資料，同時也可以快速地進行演算及做研判。因此，預防性的維護策略變得可行。
本研究將針對配備即時監控設備與電腦的多狀態衰退系統，在週期性的偵測下，提出一個預防維護策略(PM Policy). 該策略主要是在每一個適當決策時間點，規劃各可能出現狀態之最佳維護動作及進行下一次預防維護之最佳時機。其基本假設如下：
1) 系統狀態的變化採用離散時間之多狀態馬可夫鏈 (discrete time multi-state Markov chain)來描述。其轉換機率並將在每一次維護完後之瞬間改變，此後一直維持不變，直到下一次維護發生。
2) 轉移機率矩陣為上三角形以反映系統狀態的衰退。
3) 執行PM時，有多種不完全維護動作可供選用。除不做維護 (do nothing) 與置換 (replacement) 外，其他每一個維護動作皆具有失敗（未達預定目標)的風險。
4) 檢測點之間的系統運作成本隨狀態而變化。
本維護策略與其他預防保養策略不同的地方，在於每一個決策時間所規劃出之最佳維護時間與維護動作，皆是根據決策當時之系統狀態利用維護資料來制訂。因此企業可以依據系統在偵測下蒐集到之維護資料，搭配我們提出之預防保養策略，制定一個實用的預防維護策略。
本研究將以衰退過程滿足Gamma過程之系統為例，解釋如何按該維護策略執行預防維護，並比較執行前後之效益。

References
1. Anderson TW, Leo A Goodman. 1957. Statistical inference about Markov chains. The Annals of Mathematical Statistics, 28: 89-110.
2. Barlow RE, Hunter LC. 1960. Optimum Preventive Maintenance Policies. Operations Research 9, 90-100.
3. Block HW, Borges WS, Savits TH. 1988. A general age replacement model with minimal repair. Naval Research Logistics 35(5), 365–372.
4. Bagai I, Jain K. 1994. Improvement, deterioration and optimal replacement under age-replacement with minimal repair. IEEE Transactions on Reliability 43, 156-162.
5. Ben-Daya M, Alghamdi AS. 2000. On an Imperfect Preventive Maintenance Model. International Journal of Quality & Reliability Management 17(6), 661-670.
6. Bergman B. 1978. Optimal Replacement under a General Failure Model. Advances in Applied Probability 10(2), 431-451.
7. Chen A, Guo RS, Wu GS. 2000. Real-time equipment health evaluation and dynamic preventive maintenance. In Proceedings of the 9th International Symposium on Semiconductor Manufacturing 375-378.
8. Chen CT, Chen YW, Yuan J. 2003. On a dynamic preventive maintenance policy for a system under inspection. Reliability Engineering and System Safety 80, 41-47.
9. Chen D, Trivedi KS. 2002. Closed-form Analytical Results for Condition-based Maintenance. Reliability Engineering and System Safety 76, 43-51.
10. Chen M, Feldman RM. 1997. Optimal Replacement Policies with Minimal Repair and Age-dependent Costs. European Journal of Operational Research 98, 75-84.
11. Chiang JH, Yuan J. 2000. Optimal maintenance policy for a production system subject to aging and shocks. Journal of Quality Maintenance of Engineering 6, 200-216.
12. Chiang JH, Yuan J. 2001. Optimal maintenance policy for a Markovian system under periodic inspection. Reliability Engineering and System Safety 71, 165-172.
13. Cleroux R, Dubuc S, Tilquin C. 1979. The age replacement problem with minimal repair and random repair costs. Operation Research 27, 1158-1167.
14. Dekker R. 1996. Applications of maintenance optimization models: a review and analysis. Reliability Engineering and System Safety 51, 229–240.
15. Derman C. 1963. Optimal Replacement and Maintenance under Markovian Deterioration with Probability Bounds on Failure. Management Science 19, 478-481.
16. Dieulle L, Berenguer C, Grall A, Roussignol M. 2003. Sequential condition-based maintenance scheduling for a deteriorating system. European Journal of Operational Research 150, 451-461.
17. Drinkwater RW, Hastings NVJ. 1967. An Economic Replacement Model”, Operations Research Quarterly 18, 121-138.
18. Hsieh CC, Chiu KC. 2002. Optimal Maintenance Policy in a Multistate Deteriorating Standby System. European Journal of Operational Research 141, 689-698.
19. Jan AM Hontelez, Helen H Burger, Diederik JD Wijnmalen. 1996. Optimum condition-based maintenance policies for deteriorating systems with partial information. Reliability Engineering and System Safety 51, 267-274.
20. Kallen MJ, Van Noortwijk JM. 2005. Optimal maintenance decisions under imperfect inspection. Reliability Engineering and System Safety 90, 177–185.
21. Kyung S Park. 1988. Optimal continuous-wear limit Replacement under Periodic Inspections. IEEE Transactions on Reliability 37(1), 97-102.
22. Lam CT, Yeh RH. 1994. Optimal maintenance policies for deterioration systems under various maintenance strategies. IEEE Transactions on Reliability 43, 423-430.
23. Lam CT, Yeh RH. 1994. Comparison of sequential and continuous inspection strategies for deteriorating systems. Advances in Applied Probability 26, 423-435.
24. Lo ST, Wee HM, Huang. 2007. An integrate production-inventory model with imperfect production processes and Weibull distribution deterioration under inflation. International Journal Production Economics 106,493-505.
25 Marquez AC, Heguedas AS. 2002. Models for Maintenance Optimization: a Study for Repairable Systems and Finite Time Periods. Reliability Engineering and System Safety 75, 363-377.
26. Nakagawa T. 1984. Optimal policy of continuous and discrete replacement with minimal repair at failure. Naval Research Logistics Quarterly 31 (4), 543–550.
27. Nakagawa T, Osaki S. 1974. The Optimum Repair Limit Replacement Policies. Operations Research Quarterly 25, 311-317.
28. Pham H, Wang H. 2003. Optimal imperfect maintenance models. In: Pham H, editor. Handbook of reliability engineering. London: Springer 397–414.
29. Saassouh B, Dieulle L, Grall A. 2007. Online maintenance policy for a deteriorating system with random change of mode. Reliability Engineering and System Safety 92, 1677–1685
30. Singpurwalla ND, Wilson SP. 1998. Failure models indexed by two scales. Advances in Applied Probability 30, 1058–1072.
31. Sheu S, Kuo C, Nakagawa T. 1993. Extended optimal age replacement policy with minimal repair. RAIRO: Recherche Operationnelle 27(3), 337–351.
32. Sobczyk K. 1987. Stochastic models for fatigue damage of materials. Advances in Applied Probability 19, 652–673.
33. Su CT, Wu SC, Chang CC. 2000. Multiaction maintenance subject to action-dependent risk and stochastic failure. European Journal of Operational Research 125, 133-148.
34. Tahara A, Nishida T. 1975. Optimal replacement policy for minimal repair model. Journal of Operations Research Society of Japan 18, 113–124.
35. Van Noortwijk JM, Kok M, Cooke R. 1997. Optimal maintenance decisions for the sea-bed protection of the Eastern-Scheldt Barrier. In: Cooke, et al., editors. Engineering probabilistic design and maintenance for flood protection. Dordrecht: Kluwer Academic Publishers 25–51.
36. Van Noortwijk JM, Klatter HE. 1999. Optimal inspection decisions for the block mats of the Eastern-Scheldt Barrier. Reliability Engineering and System Safety 65, 203–211.
37. Wang H. 2002. A Survey of Maintenance Policies of Deteriorating Systems. European Journal of Operational Research 139, 469-489.