境外铁路普速动车组的多属性折衷选型决策

Multi attribute compromise selection decision of common speed EMUs for overseas railway projects

境外铁路建设项目中的机车车辆购置费占总投资比重大,是造价控制的重点。针对境外项目普速动车组选型决策需采用定量分析工具的要求,结合项目实际,从系统工程的角度对各备选车型方案进行排序并确定最优方案。通过运用多准则妥协解排序法(VIKOR)、改进熵权法、数据规范处理等工具,建立境外铁路普速动车组选型的决策模型,用于形成具有折衷特征的方案集。通过信息输入和决策模型使用,对马来西亚某在建客货混跑普速铁路的动车组选型问题进行算例分析,输出分别考虑群体效用最大化和个体遗憾最小化的推荐车型方案。研究结果表明:决策模型在评价指标值处理上兼顾考虑了精确数、区间数、模糊语言集合等形式,对定性和定量信息的处理方式恰当。决策模型综合考虑决策者偏好因素,实现了普速动车组选型问题的折衷处理,相较于传统成本比较法和定性分析法,具备更高可信度。敏感性分析实验结果验证了决策模型的可靠度。

The purchase cost of locomotives and rolling stocks accounts for a large proportion of the total investment in overseas railway construction projects,which is the focus of the owner's cost control.In view of the requirement that quantitative analysis tools should be used in the selection decision of general speed EMU in overseas projects,the optimal option was determined by sorting the alternative model schemes from the perspective of system engineering in combination with the actual project.Using criteria compromise solution ranking(VIKOR),improved entropy weight method,data standard processing and other tools,the decision model of overseas railway common speed EMU selection was established,which was used to form a compromise option set.By utilizing information input and decision-making model,this paper analyzed the EMU selection problem of a common speed railway under construction in Malaysia,and outputs the recommended vehicle scheme considering group utility maximization and individual regret minimization respectively.Some conclusions are drawn as follows.The decision model takes into account the forms of exact number,interval number and fuzzy language set,and the processing methods of qualitative and quantitative information are appropriate.The model comprehensively considers the preference factors of decision-makers,realizes the selection problem of common speed EMU,and verifies its effectiveness and advantages.Compared with the traditional cost comparison method,the model has higher reliability.Experimental results of sensitivity analysis verify the reliability of the decision model.