| Summary: | 碩士 === 國立宜蘭大學 === 土木工程學系碩士班 === 98 === Many uncertain factors can have a substantial effect on project planning. Factors such as weather, geographical features, technical skills, resource allocation, and construction interface, often result in ambiguous estimated project durations and precedence relationship of activities. Although traditional project scheduling may easily be applied, it is difficult to accurately assess the durations and other possible delays under the uncertainties mentioned above. As a result, accuracy and aptness are not easily achieved in project scheduling.
This research tries to establish an optimum scheduling model considering the effect of duration, cost, and feasibility at the same time. We employed fuzzy estimation to simulate the functions based on judgments and experiences associated with the manager’s estimation of project durations, and to use fuzzy number to quantify the managers’ expectation on the feasibility of each activity under different durations and precedence relationship. Next, linear programming is used to establish the computing model of the network, which takes the sum of membership function of each activity as the objective, and by optimization to deicide the durations that best meets the manager’s needs, with an aim to evaluate the feasibility of the project. Considering the construction costs in real practice, the model takes into account the trade-offs between duration and cost to find the project of highest feasibility and the lowest cost under a specific duration.
This research found that the scheduling can better meet the needs in real practice and its accuracy can be enhanced by considering the fuzziness in duration and precedence relationship of activities and the trade-offs between the time and cost. Secondly, we can improve the effectiveness of scheduling control because, under our new model, we can focus on critical paths on one hand, and prioritize different critical paths by comparing the membership function of each activity within them on the other. Thirdly, decision-makers can better determine the best scheduling plan according to their preferences by considering the time, cost, and feasibility at the same time. With this character, the proposed model can narrow the difference between the estimated duration and the actual duration often found under traditional project scheduling, which is predominantly based on cost, and better avoid the delay of project due to this estimated bias.
|
|---|
