A route efficiency analysis using Shannon entropy-based modified DEA method and route characteristics investigation for urban bus transport in India
A route efficiency analysis using Shannon entropy-based modified DEA method and route characteristics investigation for urban bus transport in India
复制标题
DOI：
10.1080/03155986.2017.1393727
复制
发表时间：
2018-01-01
期刊：
影响因子：
1.3
通讯作者：
Sengupta, Partha Pratim
中科院分区：
文献类型：
--
作者：
Adhikari, Arnab;Basu, Sumanta;Sengupta, Partha Pratim
在城市公交系统中，公交线路的效率研究、环境因素对公交线路效率的影响以及公交线路环境绩效评价等问题尚未得到足够的重视。对于绩效评估，学者们大多采用传统的数据包络分析（DEA）模型，其在处理不良参数方面的局限性限制了其在不良输出存在时的有效性。在这里，我们提出了一个基于香农熵的输出导向的综合DEA模型的整体性能测量的公交线路。除了营运及财务表现的评估外，我们亦引入一个新的“排放指标”，以评估巴士路线的环保表现。我们展示了几种改进的DEA模型的具体实施，用于测量公交线路的效率，并使用香农熵方法将不同模型的效率整合到一个分数中。我们将所提出的方法应用于加尔各答，在印度的一个城市的21条公交线路。我们的方法有利于一致的和更多的改善得分的情况下，盈利和亏损的路线，分别。根据我们的分析，路线长度，路线重叠，中间停工的数量，和成本结构的性能起着至关重要的作用。我们提供了几个特定的路线以及政策层面的建议。
In the context of urban bus transport, the efficiency study of the bus routes, the impact of the contextual factors on the efficiency, and environmental performance measurement have not been paid enough attention. For performance evaluation, scholars mostly apply conventional data envelopment analysis (DEA) model whose limitation regarding handling the undesirable parameter limits its usefulness in the presence of undesirable outputs. Here, we propose a Shannon entropy-based output-oriented consolidated DEA model for the holistic performance measurements of the bus routes. Apart from operational and financial performance measurement, we introduce a novel 'emission metric' for environmental performance assessment of the bus routes. We show the context-specific implementation of several modified DEA models for measuring the efficiencies of the bus routes and integrate the efficiencies from different models into a single score using Shannon entropy approach. We apply the proposed method on 21 bus routes of Kolkata, an urban city in India. Our method facilitates consistent and more improved score in case of profit-making and loss-making routes, respectively. According to our analysis, route length, route overlapping, number of intermediate stoppages, and cost structure play a crucial role in the performance. We provide several route-specific as well as policy-level recommendations.