Research On Learning-based Optimization Method For Repetitive Construction Projects

Repetitive project optimization problem is very common in project scheduling, Repetitive scheduling method (RSM) has several advantages over the network models in repetitive projects scheduling. On the other hand, productivity of one unit will increase if repetitive unit number or the continuous working time is increasing, and the efficiency of operationsincreasing gradually, this is called "learning effect". Activities in repetitive construction projects are repeated in each unit for the length of the jobs, which make repetitive construction projects more easily than non-repetitive projects to produce the learning effect. "Learning effect" is one of reasons which can severely affect optimization result in repetitive project scheduling. But most of the repetitive project optimization methods are not considered learning effect, which make the results are not in accordance with practical situation to a large extent. As a result, the time expended to complete the work of a unit will decrease as the number of repetitions increases.This paper aims to present an efficient method to predict the influence of the learning effect on the controlling path and the total time based on the repetitive scheduling method (RSM). Firstly,this paper describes the basic theory and concepts of repetitive projects, makes a simple analysis of optimization method for repetitive construction projects, sums up the current methods for project scheduling. The factorsinfluencing the learning rate and the currentresearch of learning effect models are summarized and the process of determining the rate presentedby a construction project. And then, we construct a nonlinear programming model for optimization method for repetitive projects based on learning effect, analyzes the critical condition for the change of the controlling points caused by the learning effect and. The effectiveness of the proposed method is demonstrated by two different types of construction projects. Finally, a new method is proposed for solving theacceleration problem with learning effect, usinga new graphical method determined activities’types. An example of concrete bridge construction project demonstrates itsvalidity. Results of this study will help project managers understand the impact of learning effect onrepetitive project, and use the learning effect to optimal scheduling scheme.