arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 1911.01282cs.CY
收藏

从Wi-Fi和蓝牙传感数据中挖掘公交客流和始发地信息

Mining Public Transit Ridership Flow and Origin-Destination Information from Wi-Fi and Bluetooth Sensing Data

Ziyuan Pu, Meixin Zhu, Zhiyong Cui, Yinhai Wang

摘要:

公交客流量和起讫点信息对于加强公交网络设计、优化公交线路和改善服务至关重要。传统的基于调查和灵巧卡数据驱动的O-D信息推断方法存在样本不足、耗时耗力、推断结果缺乏验证等问题,影响了O-D信息推断的有效性和准确性。考虑到智能移动设备在世界上的普遍性,开发了几种方法,用于通过基于预定义的阈值过滤掉非乘客MAC地址来从Wi-Fi和蓝牙感测数据估计公交乘客流量。然而,由于阈值的不确定性和缺乏定量结果验证,滤波方法的准确性仍然值得怀疑。结合乘客和非乘客特征空间重叠的假设,提出了一种基于Wi-Fi和蓝牙感知数据的公交客流量和O-D信息估计的三步数据驱动方法。所观察到的乘客流量被用作计算性能测量的基础事实。实验结果表明,该方法优于所有基线模型和现有的滤波方法。研究结果可为公交车辆管理和服务质量提升提供实时、准确的公交客流和O-D信息。

英文摘要:

Transit ridership flow and origin-destination (O-D) information is essential for enhancing transit network design, optimizing transit route and improving service. The effectiveness and preciseness of the traditional survey-based and smart card data-driven method for O-D information inference have multiple disadvantages due to the insufficient sample, the high time and energy cost, and the lack of inferring results validation. By considering the ubiquity of smart mobile devices in the world, several methods were developed for estimating the transit ridership flow from Wi-Fi and Bluetooth sensing data by filtering out the non-passenger MAC addresses based on the predefined thresholds. However, the accuracy of the filtering methods is still questionable for the indeterminate threshold values and the lack of quantitative results validation. By combining the consideration of the assumed overlapped feature space of passenger and non-passenger with the above concerns, a three steps data-driven method for estimating transit ridership flow and O-D information from Wi-Fi and Bluetooth sensing data is proposed in this paper. The observed ridership flow is used as ground truth for calculating the performance measurements. According to the results, the proposed approach outperformed all selected baseline models and existing filtering methods. The findings of this study can help to provide real-time and precise transit ridership flow and O-D information for supporting transit vehicle management and the quality of service enhancement.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑