基于PLS偏好的DEA模型在公交线路服务能力评价中的应用研究

发布时间：2018-06-08 07:51

本文选题：公交线路 + 服务能力　；参考：《北京交通大学》2017年硕士论文

【摘要】：随着中国城市化进程的加速,城市交通问题日益突出,因此政府提倡公交优先,大力发展公共交通。城市公交在一天中的通勤高峰期承担了较大的客流,满载率较高,且伴随着高峰拥堵现象,乘客常常会感觉“无车可坐”;而在平峰时段,公交客流量较小,公交线路服务能力闲置情况严重,甚至出现“空驶”现象。因此,判断公交线路在不同时段的服务能力与客流需求是否协调,是优化线路资源配置,提高线路运营效率的前提。本文以公交线路在不同时段的服务能力为研究对象,在对公交线路服务能力影响因素进行深入分析的基础上,建立服务能力PLS(Partial Least Squares)路径模型,确定模型变量,并基于海淀区约100条公交线路IC卡数据获取模型变量值。然后基于PLS路径模型分析结果,确定公交线路在平峰及高峰时段服务能力的有效影响因子和影响效应值,并以此构建公交线路服务能力评价指标体系。在对各类综合评价方法对比分析后,选择数据包络分析(Data Envelopment Analysis,DEA)法对公交线路服务能力进行评价。针对传统C2R模型(Charnes-Cooper-Rhodes Model)在无法设置指标偏好和有效决策单元不能排序这两点上的不足,本文基于PLS影响效应值确定指标偏好,将其与C2WH模型(Charnes-Cooper-Wei-Huang Model)结合,并引入理想点概念,建立了改进的DEA评价模型。最后利用改进的DEA模型对海淀上地片区的26条公交线路进行服务能力评价,验证了模型的先进性和有效性。
[Abstract]:With the acceleration of urbanization in China, urban traffic problems become increasingly prominent, so the government advocates public transport priority and vigorously develops public transport. During the rush hour of the day, the city bus takes on a large passenger flow, with a high full load rate, and with the rush congestion, passengers often feel "no car to sit on", while in the peak hours, the bus traffic is small. The service capacity of public transportation line is idle seriously, even appear "empty drive" phenomenon. Therefore, it is the premise of optimizing the allocation of line resources and improving the efficiency of line operation to judge whether the service capacity of public transportation line is coordinated with the passenger flow demand in different time periods. In this paper, the service capacity of public transport lines in different periods of time as the research object, on the basis of in-depth analysis of the factors affecting the service capacity of public transport lines, establish a service capacity partial least Squares-based path model, determine the model variables. Based on the IC card data of about 100 bus lines in Haidian District, the model variables are obtained. Then, based on the results of PLS path model analysis, the effective influence factors and effect values of bus service capacity in the peak and peak periods are determined, and the evaluation index system of bus service capacity is constructed. After comparing and analyzing all kinds of comprehensive evaluation methods, data Envelopment Analysis (DEAA) method is selected to evaluate the service capacity of public transport lines. Aiming at the shortcomings of the traditional C2R model Charnes-Cooper-Rhodes Model, which can not set the index preference and the effective decision making unit can not be sorted, this paper determines the index preference based on PLS effect value, combines it with the C2WH model Charnes-Cooper-Wei-Huang Model, and introduces the concept of ideal point. An improved DEA evaluation model is established. Finally, the improved DEA model is used to evaluate the service ability of 26 public transport lines in Shangdi area of Haidian, which verifies the advanced nature and validity of the model.
【学位授予单位】：北京交通大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：U491.17

【参考文献】