考慮界面影響之施工進度網圖模式

Abstract

施工界面（Interface）是工程界常見之一不確定（Uncertainty）因素，界面管理（Interface Management）成功與否，和工程是否能如期完工具相關關聯性這一觀念，亦被工程界廣泛接受和重視。但現有之排程技術和工具，如桿狀圖(Bar chart；Gantts chart) 、要徑法(CPM，Critical Path Method)、計畫評核術（PERT，Program Evaluation and Review Techniques）、及蒙地卡羅模擬(Monte-Carlo simulation)，均是以考慮單一作業(Activity)之工作時間長短為基礎，並以之計算整體工程之完工時間。因此對於不確定因素之考量亦多是以單一作業為範圍，而對於不同作業間同一個不確定因素所可能具有之相關性（correlation）則多未考量，如此則未能將界面此一不確定因素之特性完全表達出來。 一個具豐富經驗之排程工作者，或可於安排單一作業之工作時間時，將眾多不確定因素之影響囊括進來，但卻不能將各不確定因素之可能影響明確區分出來，更不用說是考量前後作業內同一個不確定因素之連動。而NETCOR網圖模式對瞭解作業內各不確定因素影響之大小，已提供一良好之研究基礎，對於不確定因素之相關性亦有所提及。本研究試圖以NETCOR網圖模式為基礎，並參照界面自動表達（CIAR）系統對界面之整理，建構一考量界面影響之施工進度網圖模式，此模式能表示各作業內之界面影響和前後作業內界面影響之相關性，如此即能更進一步瞭解施工進度網圖上之界面影響，並提供各階層界面管理之所需資訊。本研究最後，並以循序漸進之三個例子(2-activities，3activities，及一小網圖)分別闡述此模型之運作過程及其運用成效。 關鍵詞：界面、不確定因素、NETCOR網圖模式、界面自動表達系統

Interface is a common uncertainty factor in construction region. An idea of that interface management is related completing construction is generally accepted and respected by constructors. But existing technology of schedule and tools, for example, bar char, CPM, PERT, Monte-Carlo simulation, all base on take account of time of single activity. And by the way constructor calculate the finish time of project. So thinking of uncertain factor is also set measure to single activity. However the same uncertainty in different activities may have correlation that is always neglected. A constructor who has copious experience should envelop a mass of uncertainty factors when he evaluates time of one activity. But he must not differentiate between all uncertain factors; even more consider the time change reduces by the same uncertainty in before and after activities. However the networks model of NETCOR already provides a good base of research to understand each influence of uncertain factor in the activity. Further it also mentions the correlation of uncertainty. This dissertation is base on the networks model of NETCOR and consults the arrangement of interface by the CIAR system to establish a model for evaluating networks under interface effect. The model can express each interface effect in all activities and the relevance of both activities’ interface effect. For this reason, we can further understand the interface effect to time in schedule networks and catch the information for interface management. In the end of dissertation ,we use three examples（two activity, three activity and a small network） to explain how to use and what we can get from the model. Key words：Interface、Uncertainty、NETCOR、CIAR