摘要
根据软件项目的特点以及软件项目进度的安排,本文提出了基于贝叶斯网络的软件项目进度管理模型,在PERT图的基础上构造贝叶斯网络模型,由专家判断和工程经验确定网络中的概率参数。该模型可实现对项目进展情况的监控和控制,识别开发中对项目影响的不确定性因素,并进行反向参数学习,从而可以及时地调整不合理的开发进度,以达到优化的作用。仿真实验结果表明,该模型与实际情况相符合,应用于实际项目开发中取得了很好的效果。
According to the characteristics of software proiects, this paper puts forward a schedule management model based on the Bayesian network. This Bayesian network model is built on the basis of introducing the PERT chart. And the probability parameters are determined by the experts＇ judgment and experience. It can continually monitor and control the project, and identify the risks and uncertainty of the project, and perform reverse parameter learning. Whereby the irrational development progress is adjusted, and the role of evolutionary optimization is achieved. The simulation experiment proves that this model has good schedule predicting ability coincided with the real cases.
出处
《计算机工程与科学》
CSCD
北大核心
2011年第11期140-143,共4页
Computer Engineering & Science
基金
辽宁省教育厅基金资助项目(2009A350)
关键词
项目进度
贝叶斯网络
不确定性
管理模型
project schedule
Bayesian network
uncertainty
management model
