基于绩效预测的公共工程代建人选择研究

发布时间：2018-03-24 18:04

本文选题：代建制　切入点：代建人选择　出处：《西南交通大学》2015年硕士论文

【摘要】：选择理想的代建人不仅是代建项目最终业主(政府)最重要的决策之一,也是学术界一个具有理论意义和应用价值的热点课题。论文在对当前代建人选择研究与实践中存在的问题进行全面梳理的基础上,重点探讨了全过程企业型代建模式下复杂项目的代建人选择问题,创造性地将绩效预测的思想和方法应用于代建人选择方法、测度指标和选择路径的整体改良。基于构建的绩效预测基本分析框架,论文的主要研究内容和成果包括：(1)通过理论分析辨识了代建人项目绩效的关键驱动因素和构成要素,在运用SEM对二者之间的关系进行实证检验的基础上,确立了作为绩效预测依据的代建人能力维度、外部环境维度和代建人项目绩效的构成要素体系,并给出了基于多层次模糊数学的绩效评价方法及其实现步骤。(2)以实证分析结果和预测理论为基础,从业主选择代建人的视角构建了反映代建人项目管理能力、技术能力、经济财务能力、资源能力、外部环境和项目属性六个功能团变量与代建人项目绩效之间非线性映射关系的概念模型,提出了代建人项目绩效预测的指标体系。另外,基于BP神经网络技术设计了一套完整的预测数学模型,并给出了基于MATLAB工具箱的代建人项目绩效预测的编程实现。(3)结合国际工程咨询服务采购的实践经验,提出了基于绩效预测的代建人选择新路径：资质审查→建立动态管理的代建人名录→开标前准备→开标→资格后审→投标人项目绩效预测→面谈→面谈分数值+绩效预测分数值的综合排序→协商并确定中标人→签订代建合同。最后,提出了简单项目与复杂项目区分的十项指标及判别方法,当拟建项目属于复杂项目范畴时,基于绩效预测的代建人选择方法和路径能够为政府委托人选择到最“适合”的代建人。
[Abstract]:The choice of an ideal agent is not only one of the most important decisions of the final owner (government) of the agency project, It is also a hot topic with theoretical significance and application value in academic circles. This paper mainly discusses the problem of agent selection for complex projects under the model of full-process enterprise-oriented construction, and creatively applies the thought and method of performance prediction to the method of agent selection. Based on the basic analysis framework of performance prediction, the main research contents and achievements of this paper include: 1) the key driving factors and components of the performance of the agent project are identified by theoretical analysis. Based on the empirical test of the relationship between the two by using SEM, the author establishes the building agent ability dimension, the external environment dimension and the component elements system of the agent project performance as the basis of performance prediction. The performance evaluation method based on multi-level fuzzy mathematics and its realization steps are given. Based on the results of empirical analysis and prediction theory, the project management ability and technical capability of the agent are constructed from the perspective of the owner's selection of the agent. The concept model of nonlinear mapping relationship between the six functional group variables of economic and financial ability, resource ability, external environment and project attribute and the performance of the agent project is proposed, and the index system of the performance prediction of the agent project is put forward. Based on BP neural network technology, a set of complete prediction mathematical model is designed, and the programming realization of performance prediction of agent project based on MATLAB toolbox is given. This paper proposes a new path for agent selection based on performance forecasting: qualification review. 鈫扙stablish a directory of agents for dynamic management. 鈫扨reparation before bid opening. 鈫払id opening. 鈫扨ost-qualification trial. 鈫払idder project performance forecast. 鈫扞nterview. 鈫扖omprehensive ranking of performance Prediction score of interview score. 鈫扤egotiation and determination of winning bidder. 鈫扚inally, ten indexes and methods of distinguishing between simple project and complex project are put forward. When the proposed project belongs to the category of complex project, The method and path of agent selection based on performance forecast can select the most suitable agent for the government client.
【学位授予单位】：西南交通大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：F282

【参考文献】