| Summary: | 碩士 === 國立臺灣科技大學 === 營建工程系 === 97 === Design phase is characterized by highly coupled and interdependent activities, which result rework or iteration in this phase. In dealing with this problem, information flows have to be well organized, so it can lead to the most optimum completion time. The Dependency Structure Matrix (DSM) is a compact matrix representation that models information flows among activities in the design phase. Previous DSM concepts used basic logic of faster and fewer iteration, while assuming the number of iteration occurred to achieve optimum completion time. This thesis proposed a new methodology which combined faster and fewer iteration concepts with adding other information, i.e. Quality Equations. Quality Equations will be performed in order to reduce the lack of assumption in determining the number of iteration occurred. Resource constrained problem is also added in the thesis to make the problem close to the reality. To help coordinate/optimize the activities structure in terms of minimizing completion time, Genetic Algorithm (GA) will be conducted in this thesis. As the result, combining DSM with quality equations give more accurate expected duration since assumption regarding the number of iteration for each task can be replaced by exact number revealed from quality equations.
|
|---|
