運用多層失效樹與事件樹探討鐵路系統最佳設計

Optimal Rail System Design with the Multiple Layer Structure in Fault Tree and Event Tree

盧俊霖(Chun-Lin Lu)

指導教授 ： 賴勇成

國立臺灣大學/工學院/土木工程學研究所/碩士(2017年)

https://doi.org/10.6342/NTU201702382

摘要

鐵路系統設計屬於一項龐大的工程，下轄許多子系統，而子系統又各自包含不少的元件，且這些元件又有各自的失效頻率與成本，同時還要考量元件與元件之間的關係，因此在設計鐵路系統的過程中需要針對這些因素做權衡，決策者必須從數個不同的因子中去取捨，從中挑出最適當的系統設計。
在過去的研究當中，多將鐵路系統簡化為直接下轄子系統的架構，輔以生命週期成本或是可靠度做為考量因子。然而，鐵路系統組成事實上相當複雜且包含許多需考量因素，因此需要一套有效率且具有完善考量的決策支援方法，協助決策者找出最佳之系統設計。因此本研究研發一整合型決策支援架構以協助決策者提出最佳鐵路系統設計，其中包含失效樹轉換模組，以及最佳投資模組。轉換模組用以解決元件與元件之間的關係影響系統的失效頻率問題，能將來自失效樹分析之原始元件關係資料轉換成為可有效率處理的資料格式；而最佳投資模組則分為原始模式以及兩種延伸模式：原始模式以生命週期成本，及因元件失效所帶來後續的影響成本做為目標考量，同時加上處理串並聯及選擇關係的限制式，以求出最適當的投資。在延伸模式中，可將失效帶來的後續影響納入到模式當中做設計與考量，亦可藉由失效頻率的放大來討論對失效頻率值的不確定性影響。
在案例分析中，本研究以一旅客鐵路系統設計為主題，首先透過轉換模式前後的效率差異以證明轉換模組的必要性，並探討資料不確定性對理想情形與實際情形中系統設計之影響，以說明在實際營運時考量資料不確定的必要與優點，而設定不同的失效後續影響的要求，以說明本模式在不同要求下能針對各狀況選出最佳選擇。因此本研究所提出的決策支援模組，能夠協助決策者有彈性地根據多層架構及失效後續影響的情形，找出最適當的鐵路系統投資方案，使其能夠獲得營運上的最大回饋，並有效評估失效所帶來的後續影響以及資料不確定性所帶來的服務水準及營運狀況的差異。

關鍵字

鐵路運輸 ； 最佳系統設計 ； 失效樹分析 ； 事件樹分析 ； 資料不確定性

並列摘要

A rail system typically comprises several subsystems and corresponding components. Each component has its reliability, life cycle cost (LCC), and consequences of failure. Determining the optimal rail system design unravels different trends in these characteristics. Thus, an operator must carefully examine each alternative of the components before allocating these characteristics to achieve the optimal design. Accordingly, we develop an optimization process with two modules, namely conversion and optimal system design modules, to assist the operator in deciding the appropriate selection for a rail system design. Conversion module can transform multiple layers of Fault Tree to a simple structure and avoid nonlinear formulation in the optimization model, while the optimal system design module aims to determine the optimal investment plan for rail systems based on available alternatives. This optimization process can identify the best alternative for every subsystem according to acceptable LCC or consequences of failure. Three empirical cases were performed using all the developed models to demonstrate their applicability. The first case proves the efficiency to transform the Fault Tree structure using the conversion module. The second case illustrates that considering the data uncertainty into failure rate requires allocating additional budget to improve the delay cost under ideal and practical situations. The third case indicates that the optimization model can solve the multi-objective problem while considering LCC and consequences of failure, as well as assist the operator to decide appropriate requirement according to their demand. This comprehensive approach can help users identify the ideal balance between cost and consequences of failure to achieve an optimal rail system design.

並列關鍵字

Rail Transportation ； System design ； Fault Tree ； Event Tree ； Data Uncertainty

參考文獻

Allella, F., Chiodo, E., and Lauria, D. (2005). Optimal reliability allocation under uncertain conditions, with application to hybrid electric vehicle design. International Journal of Quality & Reliability Management, 22(6), 626-641.

Azaiez, M. N., and Bier, V. M. (2007). Optimal resource allocation for security in reliability systems. European Journal of Operational Research, 181(2), 773-786.

Burton, R. M., and Howard, G. T. (1969). Optimal system reliability for a mixed series and parallel structure. Journal of mathematical analysis and applications, 28(2), 370-382.

Chang, Y. C., Chang, K. H., and Liaw, C. S. (2009). Innovative reliability allocation using the maximal entropy ordered weighted averaging method. Computers & Industrial Engineering, 57(4), 1274-1281.

Chern, M. S., and Jan, R. H. (1986). Reliability optimization problems with multiple constraints. IEEE Transactions on Reliability, 35(4), 431-436.