TESLA: A Centralized Taxi Dispatching Approach with Global Fairness
首发时间:2019-03-25
Abstract:Taxi service plays an important part in urban transportation systems. Traditional taxi business model suffer from low efficiency. The development of smart phones and mobile computing has made centralized taxi dispatching possible. Existing taxi dispatching algorithms mostly focus on optimizing the overall profit of the entire fleet but ignore the income fairness issue of drivers. This may cause problem for driver engagement and drivers' working desire. In this paper, authors study how to improve the overall taxi drivers' revenue in a fleet while addressing the fairness issue in a central dispatching mode. They first identify the unfairness issue from the taxi dispatch process and propose a novel solution to match taxis and passengers in real time, namely cenTralizEd diSpatch with gLobal fAirness (TESLA). They design a dispatching system to improve driver's revenue efficiency and reduce passengers' waiting time under the premise of ensuring driver income fairness. The experimental results show that their TESLA approach outperforms the real taxi operation strategy and a baseline approach.
keywords: Software engineering Centralized dispatching Global fairness Revenue efficiency
点击查看论文中文信息
TESLA:一个融入全局公平性的出租车集中调度方法
摘要:出租车服务在城市交通系统中扮演着一个重要的角色,然而,传统的出租车商业模式一直存在效率低下的问题。智能手机和移动计算的发展使出租车的集中调度成为可能,现有的出租车调度算法主要集中在优化系统的的整体利润,但忽略了司机的收入公平问题,这可能会导致出租车司机的工作积极性降低。在本文中,作者研究如何在集中调度模式下提高出租车司机的整体收入,同时解决出租车司机之间的收入公平性问题,他们首先从出租车调度过程中发现不公平问题,并提出一种新的解决方案(TESLA),来实时地匹配出租车和乘客。他们设计了一个调度系统,在保证司机收入公平的前提下,提高出租车司机的收入效率,减少乘客的等待时间,并且通过实验结果表明了他们的TESLA方法优于真实的出租车运营策略和基线方法。
基金:
引用
No.****
同行评议
勘误表
TESLA:一个融入全局公平性的出租车集中调度方法
评论
全部评论0/1000