在建築工程施工過程中,由於人、事物、環境與管理上的不確定性,造成施工風險不確定性尤甚於其他產業,故理應投注較多心力於工程風險的管理與控制。但國內大多數的營造廠商對於風險管理概念認知極低,甚至對風險採取不予理會的心態,假若營造廠商能於事前對工程風險進行風險評估分析,則可將人員傷亡或財務損失等之風險降至最低。 因此,本論文將透過風險管理與評估相關文獻的分析與整理,藉此建構建築工程施工風險評估概念基礎架構,再導入模糊理論與多準則決策概念作為推估風險發生可能性與計算災害風險程度之方法與技巧。結合上述架構與原理,利用具備物件導向功能之C#軟體,開發一套建築工程施工風險評估系統平台。最後再透過遺傳演算法來處理綜合評估嚴重程度最佳化問題,進而減少使用人員之間對於事物判斷上的差異,使分析結果更趨合理化,希冀能使整套評估系統更加完善。
Because of the uncertainty of the employees, equipments, environment and management during construction stages, the construction risk is more uncertain than other industries. So we should take care in the project risk management and control. But most constructers in Taiwan don’t have the sense of risk management, or ignore it. If the constructers can take risk assessment before construction, casualties or financial loss should be reduced to a minimum. This thesis collects and analysis the relevant literature of risk management to establish a framework of risk assessment in building construction. In which, the fuzzy theory and multiple criteria decision making concept are integrated in estimating the risk possibility and calculating the disaster risk. Furthermore, using C # software with object-oriented utility, the thesis develops a risk assessment system platform. Finally, the genetic algorithm was used to deal with the synthetic conclusims of many users’ opinion.As a result, an optimum risk assessment of construction engineering is able to be established.