5. 選擇適切的分析方法
本帖最後由 hlperng 於 2017-1-8 16:26 編輯5. Selecting the Appropriate Analysis Method
選擇適切的分析方法
Selecting methods to implement into a dependability programme is a highly individualized process, so much so that a general suggestion for a selection of one or more of the specific methods cannot be made. The selection of appropriate methods should be carried out by a joint effort of experts from the dependability and system engineering field. Selection should be made early in the programme development and should be reviewed for applicability.
Selecting methods can be made easier, however, by using the following criteria:
a) System complexity: complex systems, e.g., involving redundancy or diversity features, usually demand a deeper level of analysis than simpler systems.
a) 系統複雜性：
b) System novelty: a completely new system design may require a more thorough level of analysis than a well-proven design.
b) 系統新奇性：
c) Qualitative versus quantitative analysis: is a quantitative analysis necessary?
c) 定性或定量分析：
d) Single versus multiple faults: are effects arising from combination of faults relevant or can then be neglected?
d) 單一或多重故障：
e) Time or sequence-dependent behaviour: does the sequence of events play a role in the analysis (e.g., the system fails only if event A is preceded by B, not vice versa) or does the system exhibit time-dependent behaviour (e.g., degraded modes of operation after failure, phased missions)?
e) 時間或次序相關：
f) Can be used for dependent events: are the failure or repair characteristics of an individual item dependent on the state of the system?
f) 是否可用於相關事件：
g) Bottom-up versus top-down analysis: usually bottom-up methods can be applied in a more straightforward manner, while top-down methods need more thought the creativity and may therefore be more error-prone.
g) 由下而上或由上而下分析：
h) Allocation of reliability requirements: should the method be capable of quantitative allocation of reliability requirements?
h) 可靠度要求配當：
i) Mastery required: what level of education or experience is required in order to meaningfully and correctly apply the method?
i) 所需的專業：
j) Acceptance and commonality: is the method commonly accepted, e.g., by a regulatory authority or a customer?
j) 接收性與共用性：
k) Need for tools support: does the method need (computer) tool support or can it also be performed manually?
K) 是否需要工具支援：
l) Plausibility checks: is it easy to inspect the plausibility of the results manually? If not, are the tools available validated?
l) 合理性檢查：
m) Availability of tools: are tools available either in-house or commercially? Do these tools have a common interface with other analysis tools so the results may be re-used or exported?
m) 工具可用性：
n) Standardization: is there a standard which describes the feature of the method and the presentation of results (e.g., symbols)?
n) 標準化：
Table 2 gives an overview of various dependability analysis methods and their characteristics and features. More than one method may be required to provide a complete analysis of a system.
表2：精選可恃性分析方法之特性
方法適合複雜系統適合新系統設計定量分析適合故障組合適合處理後果相關性可用於相關事件由上而下或由下而上適用於可恃性配當必要克服(從低到高)接受性與通用性需要輔助工具合理性檢查工具可用性IEC標準
失效率預估否是是否否否由下而上是低高平均是高61709
故障樹分析(FTA)是是是是否否由上而下是平均高平均是高61025
事件樹分析(ETA)不建議不建議是不建議是是由下而上不建議高平均平均是平均
可靠度方塊圖分析(RBD)不建議不建議是是否否由上而下是低平均平均是平均61078
馬可夫分析 是是是是是是由上而下是高平均高否平均61165
派萃網絡分析 是是是是是是由上而下是高低高否低
失效模式與效應分析(FMEA)
不建議不建議是否否否由下而上不建議低高低是高60812
危害與操作分析(HAZOP)是是否否否否由下而上否低平均低是平均61882
人員可靠度分析是是是是是是由下而上否高高平均是平均
應力強度分析
不適用不適用是不適用不適用否不適用否高平均高是平均
真值表
否是是是否否不適用是高平均高否低
統計可靠度方法
是是是是是是不適用不建議高平均高平均低60300-3-5
NR：「不建議」，可用於簡單系統，不建議單獨使用此方法，必須與其方法一起使用。TD：由上而下 (Top-down)
BU：由下而上 (Bottom-up)
Avg：平均 (Average)
NA：「不適用」，該項準則不適用於此一方法。
頁:
[1]