• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>International Symposium on Transportation and Traffic Theory; 20050719-21; University of Maryland, College Park,MD(US) >RATIONALITY AND HETEROGENEITY IN TAXI DRIVER DECISIONS: APPLICATION OF A STOCHASTIC-PROCESS MODEL OF TAXI BEHAVIOR
【24h】

RATIONALITY AND HETEROGENEITY IN TAXI DRIVER DECISIONS: APPLICATION OF A STOCHASTIC-PROCESS MODEL OF TAXI BEHAVIOR

机译：出租车驾驶员决策中的合理性和异质性：出租车行为随机过程模型的应用

获取原文
获取原文并翻译 | 示例
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

A Markov process model of taxi behavior has been proposed in this study. Although passenger demand is treated as inelastic, the model system has offered several important insights. In particular, it has demonstrated that the taxi system, just like traffic flow on road networks, contains a discrepancy between the social optimum and user equilibrium; when individual taxi drivers operate while being motivated to maximize their own individual gains, the system as a whole settles at a suboptimal point. This is a case of social dilemma, i.e., as a result of individuals' pursuit of their own gains, the total gain of the system as a whole declines, and eventually the individuals' gains decrease below the level that can be attained if they cooperated toward a social optimum. Kitamura et al. (1999) illustrated this for the case of public transit ridership based on a model of commute travel in an abstract city; if commuters can agree on a rule to restrict their car use, then every commuter's gain can improve. The presence of social dilemmas in taxi operation implies that improved system performance must be based on some modification of driver behavior. The study results suggest that drivers' inclinations toward cruising to pick up passengers must be restrained by rules, regulations, or incentives. By estimating a random-effect discrete choice model of drivers' choice between cruising and waiting at a taxi bay, the study has shown that taxi drivers behave rationally, motivated by the reduction of excess time. Furthermore, they are biased against waiting at a taxi bay, and react more strongly to an increase in the excess time involved in waiting than to that involved in cruising. This propensity toward cruising contributes to the above discrepancy between social optimum and equilibrium because the system must have a much larger fraction of taxis waiting at taxi bays to attain a social optimum. The empirical analysis of driver behavior has also shown that drivers are heterogeneous and their preferences toward cruising or waiting vary from driver to driver; some drivers consistently prefer to cruise, or to wait at a taxi bay, than others. It has been indicated that cruising becomes more attractive when it involves less uncertainty, while waiting becomes more attractive when it is more uncertain. The study results suggest certain directions toward the improvement of taxi operation in large cities. Yet the validity and generality of the study results must be rigorously examined and the model must be extended to represent taxi operation more realistically. In particular, more rigorous treatment of time of day in the choice model is desired. Also desired is representing the effects of other drivers' choices in the discrete choice model formulation. Subjects for future research also include: testing the validity of the Markovian assumptions-especially history independence and negative-exponential and Erlang sojourn durations; incorporating elastic passenger demand and representing supply-demand equilibrium; examining the nature of taxi drivers' attitude toward risk; and developing efficient procedures to simulate the system with heterogeneous transition probabilities that account for heterogeneity as well as the spatial dimension.
机译：这项研究提出了出租车行为的马尔可夫过程模型。尽管旅客需求被视为缺乏弹性，但模型系统提供了一些重要的见解。特别是，它已经表明，出租车系统就像公路网络上的交通流量一样，在社会最优与用户均衡之间也存在差异。当个体出租车司机在积极发挥自身最大利益的同时进行操作时，整个系统将处于次优状态。这是社会两难的情况，即，由于个人追求自己的收益，整个系统的总收益下降，最终个体的收益下降到低于他们合作所能达到的水平走向社会最优。北村等。 （1999）基于抽象城市的通勤旅行模型，说明了公共交通出行情况。如果通勤者可以就限制使用汽车的规则达成共识，那么每个通勤者的收益都会提高。滑行操作中存在社会困境，这意味着改进的系统性能必须基于驾驶员行为的某种修改。研究结果表明，驾驶员倾向于乘车巡航要受到规则，法规或奖励措施的约束。通过估计驾驶员在出租车和在出租车停靠之间的选择之间的随机影响离散选择模型，研究表明，出租车司机的行为理性，受多余时间减少的影响。此外，他们偏向于在出租车站等候，并且对等待所涉及的额外时间增加的反应比对巡航所涉及的时间要强烈。这种巡航的倾向加剧了社会最优与均衡之间的上述差异，因为该系统必须有很大一部分出租车在出租车站等候才能达到社会最优。对驾驶员行为的实证分析还表明，驾驶员是异质的，他们对巡航或等待的偏好因驾驶员而异。一些驾驶员始终比其他驾驶员更喜欢巡航或在出租车站等候。已经表明，巡航在涉及较少不确定性时变得更具吸引力，而等待在涉及不确定性时变得更具吸引力。研究结果为改善大城市的出租车运营提供了一定的方向。然而，必须严格检查研究结果的有效性和一般性，并且必须扩展模型以更实际地代表出租车的运营。特别地，期望在选择模型中对一天中的时间进行更严格的处理。还需要在离散选择模型公式中表示其他驾驶员选择的影响。未来研究的主题还包括：测试马尔可夫假设的有效性-特别是历史独立性以及负指数和Erlang逗留时间；纳入弹性的客运需求并代表供求平衡；检查出租车司机对风险态度的性质；并开发有效的程序来模拟具有异质转移概率的系统，该异质转移概率考虑了异质性以及空间维度。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号