| Title |
Non-normality of Probability Distribution of Project Completion Times in Simulation based Scheduling |
| Keywords |
Simulation ; Scheduling ; Stochastic Probability Theory ; Project Completion Time ; PERT ; MATLAB |
| Abstract |
Existing PERT and simulation based scheduling methods have been based on the assumption that the simulation output data, Project Completion Times (PCTs), follow normal distribution. However, this paper verifies that the normality assumption of Distribution of PCTs in Simulation based Scheduling is not always acceptable and the existing belief may lead to misleading results. The assumption has been well accepted in construction scheduling community. However, substituting the quantification method that measures the effect caused by the assumption may improve the reliability of the simulation method. This paper presents a risk quantification method relative to the probability distribution of PCTs. The method is implemented as an algorithm in MATLAB. Several series of simulation experiments have been carried out to analyze a set of simulation output data which are obtained from different type of Probability Distribution Function (PDF) assigned to activities’duration in a network. The method facilitates to find the effect of PDF type and its parameters. This article describes the procedure necessary for performing the risk quantification method along with the findings. This paper contributes to improving the reliability of simulation based scheduling method, as well as increasing the accuracy of analysis results. |