應用簡化之CPM/PERT模式預估專案工期之研究

Abstract

營建產業於目前台灣的產業市場日趨重要，如何讓營建工程於期限內順利完工，是很重要的課題，然而在實務上較廣泛被使用的進度規劃技術，如桿狀圖或是要徑法，這些確定性的進度規劃技術在面對較大規模的工程專案時，對於要徑的計算上缺乏效率；因此電腦模擬的技術可以幫助快速、正確地求算工期。Lu與AbouRizk於2000年提出簡化CPM/PERT模式之模擬模型，其前推式的計算與CYCLONE類似；但逆推式的計算採用較簡單且較有效率的求解方式，不需求出各節點的總浮時便能找到總工程之要徑。另外此模式具有要徑指數（Criticality Index，簡稱CI）的功能，能夠算出每項作業在工程專案中，成為要徑的機率，較傳統的要徑檢查更具完整性，對管理者也更具參考價值。 本研究利用Lu與AbouRizk（2000）之簡化要徑法/計畫評核術（CPM/PERT）模擬模式，應用於個案工程專案中，取代傳統的CPM/PERT模式，幫助進度管理者預估較確定性的工期時間，並且計算網路中各路徑成為要徑的機率。另外經由與建築設計師及施工單位的訪談，探討工程中各作業工項變異的可能因素，最後在趕工情境之下進行模擬，比較個案工程專案於不同情境之下，要徑的變異性。 個案工程專案的模擬結果顯示，簡化CPM/PERT模式同樣具有預估工期的能力，且能求得較確定性的完工工期、對於要徑的判斷，能提供更完整的工程資訊；而在專案趕工的情境之下，Lu與AbouRizk（2000）之簡化CPM/PERT模式能判斷6種不同的要徑情況，提供較全面的作業要徑機率資訊，幫助管理者進行風險分析，增加在工期預估的效率與準確性。

It is quite important to finish a project or a construction in the deadline for the project managements. Bar chart and classic critical path method (CPM) are both deterministic tools which is easy applying to the project scheduling. However, lack of efficiency to compute the critical path of larger scale project limit the effectiveness. Construction simulation by computer can provide prompt and accurate schedule information, such as critical path and the total project time. Lu and AbouRizk (2000) presented a simplified CPM/PERT Simulation Model, which has been done to overcome the limitations and enhances the computing efficiency of classic CPM/PERT model. This simplified model incorporates a simplified critical path identification method by “Critical Index, CI,” and supplies more comprehensive activities information than classic CPM/PERT model. This new solution to CPM network, presented by Lu and AbouRizk (2000), can offer project management with a convenient tool to assess alternative scenarios based on computer simulation and risk analysis.

This study uses a practical case to verify the application of the simplified CPM/PERT simulation model. It validates the simplified model and compares it to classic CPM/PERT analysis. The result shows that, in the basic scenario, the simplified model can provide more completed critical path information than classic model. Moreover the simplified model can judge six different critical path conditions in the crashing scenario. It can improve CPM/PERT model analysis and reinforce the accuracy and efficiency on project prediction.