考量風險偏好與有限資源限制下之專案風險決策模式

Abstract

專案執行的過程中，往往會遇到許多意外事故導致專案時程延宕或金錢損失，這些事故稱之為專案風險。因此，為降低專案風險所帶來的影響，有效的專案風險管理模式實為重要之議題。 本研究整合過去國內外學者所提出之專案風險分析模式，並引入風險偏好的概念，建構一專案風險決策模式，使決策者可依據自身之風險偏好，針對各項風險回應策略作評估，並在有限資源條件限制下，進行決策分析。本研究所建構之專案風險決策模式乃結合「專案風險失效模式與效應分析」(project Risk Failure Mode and Effects Analysis，簡稱RFMEA)與「專案風險管理回應策略選擇模式」，以風險降低值取代效益作為風險回應策略之評估指標，並應用「以效用為選擇基礎之決策支援問題」(utility based-selection decision support problem，簡稱u-sDSP)模式以評估決策者對於各項風險回應策略之效用值，最後使用整數規劃的原理，在滿足決策者效用最大化與有限資源條件限制的前提下，選出一組風險回應策略組合。而此組合，即為對此決策者而言之最佳風險回應策略組合。 由於專案風險管理回應策略選擇模式乃是使用效益作為風險回應策略的評估指標，但本研究發現此指標有其不當之處，因此本研究乃使用風險降低值作為評估指標。最後以一魚雷開發專案為例，套用專案風險管理回應策略選擇模式與本研究所建構之專案風險決策模式進行分析，所得之結果顯示，本研究所得之結果要優於專案風險管理回應策略選擇模式所得之結果，且本研究建構之專案風險決策模式亦考量到決策者風險偏好之問題，因此在實際應用上，更能得到符合決策者心中理想之決策。

During the execution of a project, unexpected events usually cause delay and financial loss, these are termed project risks. As a result, in order to reduce the impact of project risk, effective risk management model becomes an important issue. This study integrates project risk analysis models proposed by scholars and introduces the concept of risk preference to establish a project risk decision model. Decision makers can evaluate each risk treatment strategy according to their own risk preference. This study establishes a project risk decision model and uses “risk decrease” to replace “effect” as the evaluating index. Furthermore, the utility based-selection decision support problem model (u-sDSP) is employed to evaluate the utility of risk treatment strategies for decision makers. Finally, the principle of integer programming is adopted to determine the optimal set of risk treatment strategies which maximizes effects with limited resources. Taking a case of Torpedo Development Project as an example, the proposed combined model is employed in the analysis, and the result is compared with those obtained form other models. In conclusion, results obtained from the proposed model of this study is better. The reason is attributed to that the project risk decision model established by this study has taken risk preference into consideration. Thus, in practical application, the proposed model can provide better decision making solutions for decision makers.