現今以精密設備為主的生產線需要依賴高可靠度與高穩定性的設備，以避免突發故障所造成的損失。為維持設備在良好的運作狀態，則必須透過設備自動化與預防維護措施來達成。因此如何利用感應器來即時偵測設備運作狀態、並利用電腦中心根據即時資訊來制定有效的動態預防維護策略，為當今最重要的研究主題之一。
本研究即針對具有老化及衰退特性之設備，利用設備健康指標（Health Index）將設備多元的偵測參數整合，判斷出設備即時運作狀態。此運作狀態轉換以一離散時間馬可夫衰退模式來描述。本研究以（1）上三角轉換矩陣描述設備衰退；（2）更以老化因子（aging factor）描述此轉換矩陣因設備老化現象而衰退（加速）的現象；（3）並依據設備的運轉資料估計模式相關參數（老化因子及設備初始轉換矩陣）；（4）假設設備在各狀態有多種不完全維護動作可選擇，且維護動作具有風險性。

本研究提供此模式相關參數的估計方法。且為了使模式更符合實際狀況之應用，加入了維護動作時間的考量，在有限的預防維護策略期數範圍內，以每個循環的期望單位時間總維護成本最低之原則，制訂設備各狀態的最佳維護策略。

The advanced technology industries require high precision and highly stable equipment to avoid the cost caused by failure occurrence. Equipment automation and preventive maintenance must be executed to keep equipment within proper operating states. It has always been an important topic for study in inspecting the operating states of equipment by sensors and determing the dynamic preventive maintenance policies according to the real-time condition by computer analyzing centers.
This research is focused on the deteriorating system. The real-time equipment operating state can be generated by using health index integrating the parameters inspected by sensors. In this research, a discrete time Markovian multi-state deteriorating model is adopted to describe the state transition of equipment. The model also includes (1) the deteriorating of operating states ( via upper triangular probability transition matrices) and (2) the aging of transition probabilities ( via aging factor )；(3) Base on historical data, we can estimate the model parameters （aging factor and initial transition matrix） and (4) assume that multiple imperfect actions with risk are available.

The estimation methodologies of the model parameters are also provided in this research. For the practical application of the model, the maintenance time is also taken into consideration. According to the real-time equipment status, the optimal maintenance policy of each operating state can be generated by the rule of the minimum expected total cost per unit time for each cycle during finite time interval in the near future.

[1] Barlow, R.E., Hunter, L.C., "Optimum preventive maintenance policies.",
Operations Research 8, 90-100, 1960
[2] Berg, M., Epstein, B., " A modified block replacement policy.", Naval
Research Logistics 23, 15–24. (1976)
[3] Block, H.W., Langberg, N.A., Savits, T.H., "Repair replacement policement
policies." , Journal of Applied Probability 30 (1), 194-206, 1993
[4] Cakanyildirim, M.; Roundy, R.O, ”Evaluation of capacity planning practices
for the semiconductor industry”, IEEE TRANSACTIONS ON
SEMICONDUCTOR MANUFACTURING, v15, (3) (2002)
[5] Canfield, R.V., "Cost Optimization of Periodic Preventive Maintenance", IEEE
Transactions on Reliability, Vol.R-35, No.1, 1986
[6] Chen, A., Guo, R.-S., Wu, G.-S., "Real-time equipment health evaluation and
dynamic preventive maintenance.", The 9th International Symposium on
Semiconductor Manufacturing, Tokyo, Japan, September, 2000.
[7] Chiang, J.H., and Yuan, J., “Optimal maintenance policy for a Markovian
system under periodic inspection”, Reliability Engineering and System Safety
71 (2001) 165-172.
[8] Chun, Y. H.,"Optimal Number of Periodic Preventive Maintenance Operations
under Warranty", Reliability Engineering and System Safety 37, 223-225,
(1992)
[9] Hongzhou Wang.,"A survey of maintenance policies of deteriorating systems.",
European Journal of Operational Research 139 (2002) 469-489
[10] Jardine, A.D.S., Banjevic, D., and Makis, V., “Optimal Replacement Policy and
the Structure of Software for Condition-based Maintenance” , Journal of
Quality in Maintenance Engineering 3 (1997) 109-119
[11] Kercecioglu, D., “Maintainability, Availability, & Operational Readiness
Engineering”, Prentice Hall PTR, 1995
[12] Lam, C.T., and Yeh, R.H., “Optimal Maintenance-Policies For Deterioration
Systems Under Various Maintenance Strategies”, IEEE Transactions on
Reliability 43 (1994) 423-430.
[13] Liu, X.,Makis, V.,Jardine,A.K.S.,"A replacement model with overhauls and
repairs.",Naval Research Logistics 42,1063-1079, 1995
[14] Nakagawa, T., "Optimal policy of continuous and discrete replacement with
minimal repair at failure.", Naval Research Logistics Quarterly 31 (4),
543-550, 1984
[15] Nakagawa,T.,"Periodic and sequential preventive maintenance policies.", Journal of Applied Probability 23 (2),536-542, 1986
[16] Nakagawa, T.," A summary of periodic replacement with minimal repair at
failure." ,Journal of the Operations Research Society of Japan 24, 213–228.
(1981a)
[17] Nakagawa, T., " Modified periodic replacement with minimal repair at
failure.", IEEE Transactions on Reliability R 30, 165–168. (1981b)
[18] Peter Van Zant, “Microchip Fabrication” 4th edited (2001)
[19] Pham, H. and Wang, H., “Imperfect Maintenance”, European Journal of
Operational Research, 94, pp. 425-438, 1996
[20] Sheu, S., Kuo, C., Nakagawa, T., "Extended optimal age replacement policy
with minimal repair.", RAIRO: Recherche Operationnelle 27 (3), 337-351,
1993
[21] Su, C.T., Wu, S.C., and Chang, C.C., “Multiaction maintenance subject to
action-dependent risk and stochastic failure”, European Journal of Operational
Research 125 (2000) 133-148.
[22] Tahara, A., Nishida, T.,"Optimal replacement policy for minimal repair
model.",Journal of Operations Research Society of Japan 18 (3-4), 113-124
(1978)
[23] Tango, T.,"Extended block replacement policy with used items.",Journal of
Applied Probability 15, 560-572 (1978)
[24] Wang, H.,Pham, H., "Some maintenance models and availability with
imperfect maintenance in production systems.", Annals of Operations
Research 91,305-318, 1999
[25] W.J. Hopp.,S.C. Wu.,"Machine Maintenance with Multiple Maintenance
Action.", IIE Transactions,22, 226-233, (1990)
[26] Wood, A.P., “Optimal Maintenance Policies for Constantly Monitored System”, Naval Research Logistics 35 (1988) 461-471
[27] 吳耿碩,陳正剛,郭瑞祥,「及時機台健康狀況以及動態預防策略」,國立台灣
大學工業工程學研究所碩士論文1999.
[28] 陳怡文, 『多種保養動作下動態預防保養策略之研究』,國立清華大學工業
工程與工程管理學系碩士論文2000
[29] 郭瑞祥,陳正剛,曾傳蘆,馮蟻剛,楊斌虎,李俊霖,吳慶鴻,林雪蘭,黃順傑,李益
青,張聖規,李明益, “A real-time equipment monitoring and fault detection
system”, Semiconductor Manufacturing Technology Workshop (1998) 111-121.