Academic Journal of Engineering and Technology Science, 2021, 4(1); doi: 10.25236/AJETS.2021.040106.

## Project Completion Probability Analysis Based on Bayesian Network

Author(s)

Chunli Qiao

Corresponding Author:
Chunli Qiao
Affiliation(s)

College of Magement, Shanghai University, Shanghai 200444, China

### Abstract

In the case of not considering the risk,  the calculation of the completion probability of the project by Bayesian network will exaggerate the calculation result. Based on this problem,  the calculation method of the completion probability of a project affected by risks is proposed. This paper first analyzes the differences between Bayesian network considering risks or not,  and establishes a new Bayesian network model based on the advantages of Bayesian network calculating probability and the risk situation encountered in construction. Then,  it introduces the calculation method of completion probability considering risk,  which mainly includes three types: one is only affected by one of the predecessor activities and risks: the other is affected by both,  but the predecessor activities and the current work are not affected by the same risks; the third is the two work are affected by the same risk factors. Calculate the completion probability of the work until the last work based on the above method. Finally,  Combined the model with an example to verify its effectiveness. By comparing the completion probability considering risks with that not considering risks,  we can see that due to the uncertainty of the environment,  the model considering risks has a wider application in the project. It can calculate the completion probability according to the real-time situation and help constructors take measures to ensure its completion in time,  which also provides an effective decision basis for project constructors and project builders.

### Keywords

Bayesian network, probability of completion, risk, predecessor activities

### Cite This Paper

Chunli Qiao. Project Completion Probability Analysis Based on Bayesian Network. Academic Journal of Engineering and Technology Science (2021) Vol. 4 Issue 1: 62-69. https://doi.org/10.25236/AJETS.2021.040106.