基于数据包络分析的隧洞施工仿真多方案评价与优选

Project evaluation and optimization of tunnel construction using DEA simulation and analysis

针对以往隧洞施工仿真多方案优选采用基于特尔菲法用线性加权求解的多目标规划存在主观认识依赖程度高、加权系数缺少依据、缺乏后分析等问题,本文建立了耦合进度和成本的隧洞施工仿真数据包络分析(DEA,Data Envelopment Analysis)多目标综合评价模型,结合工程实例,对隧洞施工方案进行了综合评价与优选。在评价中,采用DEA模型分析了不同方案的相对有效性,方案的优劣顺序,使隧洞施工仿真多方案优选有了统一的评价准则,运用投影定理获得了其他非有效方案失效原因,并与多目标规划法进行了对比,结果表明DEA方法便于决策者对施工方案进行调整和分析,为决策者提供了科学的依据。

We developed a DEA multi-objective evaluation model of tunnel construction simulation to couple project schedule and cost and overcome the shortcomings of traditional evaluation based on multi-objective programming, i.e. highly dependent on subjective cognition, difficult to define weighted coefficient and short of post analysis. This new model was applied to analysis on the relative efficiency of different practical projects, the rank of project superiority, and the failure causes of non-effective projects. The results show that, in comparison to the traditional method, the DEA method gives objective evaluations and provides optimization suggestions for project design.