工程进度计划中的知识挖掘与应用

建设企业在工程实践中积累着大量历史项目的进度计划及管理文档资料,其中蕴含有丰富的工程经验和知识。受制于技术手段的不足,这些知识无法得到有效的识别和利用。该文基于图挖掘的理论,提出了一种能够从进度计划的网络图和Gantt图中识别出具有固定组合模式且频繁出现的工序集合的算法。试验测试和应用的结果表明,算法具有良好的可行性和有效性。算法的提出有助于项目管理人员从已有的进度计划中识别出有价值的信息和知识,为当前工程项目的计划制定和进度控制提供决策依据,以及为研究和开发工程进度管理和决策支持系统提供技术基础。

Knowledge discovery in engineering schedules and its applications

Many historical schedule files are stored as archives in engineering practices with much valuable information(e.g.,experience and lessons) hiden in the files.However,due to the limitations of technology approaches,such information cannot be effectively relearned and utilized.This paper presents an algorithm for recognizing the working procedures that frequently appear in a fixed pattern from Gantt charts and network schedules.This algorithm can help mine information and knowledge from existing schedules and support decision-making in generating,reviewing,and modifying engineering schedules.The results of experiment and case study show that the algorithm is very effective and can facilitate the reuse of historical schedules and the development of new schedule management and decision support systems.