模糊失誤樹分析在品質管制之應用

QUALITY CONTROL USING FUZZY FAULT TREE ANALYSIS

失誤樹分析是一種廣為國外系統安全與工業製程品管界認同，可有效地對失誤事件之因果關係，同時進行定性與定量分析的技術。由於造成品質缺失的因子中，例如人為錯誤，多屬含糊與不精確的事件且實務上係以主觀的語意評估，故難以準確地估計出缺失發生的機率，此為失誤樹應用上最主要的限制。鑑此，本研究提出一模糊失誤樹分析模式，用以評估模糊事件的失誤率；並以高雄市建築物混凝土樓版發生龜裂的評量為案例，說明本模式的分析流程與潛在的適用性。結果顯示，本模式的分析結果與高雄市工程品質查核的結果差異不大。因此，本模式可作為一有用的品質管制技術，用以協助品管人員有效地估計品質的缺失率。

Fault tree analysis (FTA), a widely recognized technique for system safety and process quality control, enables to effectively and simultaneously evaluate causes-and-effect relationship in qualitative and quantitative manners. Factors causing quality defects such as human errors are vague and imprecise events which are linguistically assessed by experts’ subjective judgment tend not be able to accurately estimate fault probabilities. To overcome this disadvantage, a fuzzy fault tree analysis model is proposed. A case study of assessing overall crack rate of concrete slab of the building in Kaohsiung city is performed to illustrate the potential aptness of this model. The results show that the fault rates estimated by the model are close to those obtained from the actual audit outcomes, indicating this model a useful tool that can assist quality engineers to effectively assess fault rate.