다양한 학습곡선을 반영하는 시그모이드 함수 기반 진도관리 시뮬레이션 모델

Abstract

Sigmoid modeling method, one of the widely used learning curve modeling methods, has its limits in implementing construction project cash flows, because it generates learning curve with just one single complicated formula. Therefore it needs to be developed to cope with practical situations - where many factors affect the shape of learning curves. This study adopts system dynamics modeling method to simulate S-shaped learning curves. The simulation model was constructed to apply various factors in modeling learning curves. It introduced several factors such as initial delay variance, cost variance, target date, productivity variance and these factors enable the simulation model to apply various situations of construction projects, such as schedule delay, cost increase, lagging work speed. While conventional sigmoid curve modelling is difficult to reflect changes in the middle of the project, the proposed model allows variable adjustment any time of the project progression. Statistical evaluation showed that it is 955 confident that the simulated result of the proposed model matches conventional sigmoid curve. It can be used for finding appropriate daily productivity by comparing the learning curves of target duration and actual duration, and can also be helpful forecasting cash flow features for S-shaped learning curve model.