基于蓝牙技术的城市道路短时旅行时间预测方法研究

发布时间：2018-05-30 15:14

本文选题：旅行时间 + 蓝牙技术　；参考：《上海交通大学》2015年硕士论文

【摘要】：道路的旅行时间是反映交通状态的重要指标之一,也是智能交通系统中研究的重要内容之一。道路旅行时间的准确预测是现代交通诱导系统和先进的出行者信息系统的重要基础。交通数据的采集是进行道路旅行时间预测的前提,传统的交通数据采集方式主要有基于感应线圈、基于车牌识别和基于GPS浮动车,过去大多数的关于道路旅行时间预测的研究也主要基于这三种方式所采集的交通数据来建立模型对未来的道路旅行时间进行预测。随着技术的发展,尤其是智能手机的普及、车载信息系统的完善以及蓝牙技术的不断进步,利用蓝牙技术实时采集道路交通数据引起了关注并在一些城市得到了实际应用。该项交通数据采集技术主要是通过检测道路上车辆中的智能手机、车载蓝牙或者蓝牙耳机等电子设备中的蓝牙信号,并获取这些电子设备中的蓝牙MAC地址,通过MAC地址的匹配来实时获取道路的旅行时间等交通信息。本文详细介绍了蓝牙技术实时采集交通数据的原理和基于蓝牙技术的道路旅行时间数据筛选以及平均旅行时间的估计方法。通过问卷调研了解蓝牙设备在实际中的应用情况,估计出车辆中的蓝牙信号被检测到的比例约为2.7%到4.3%之间。为了对未来的道路旅行时间进行预测,本文基于卡尔曼滤波理论和支持向量机理论以及粒子群优化算法,分别建立了基于卡尔曼滤波的短时旅行时间预测模型和基于支持向量机的道路旅行时间预测模型。在上海市闵行区沪闵公路上,利用蓝牙技术所采集的实际的道路旅行时间数据为例,对所建立的预测模型的进行了参数选择、样本数据集的训练和预测求解,并与BP神经网络预测模型进行了对比分析,结果表明本文所提出的预测模型具有一定的有效性。
[Abstract]:The travel time of the road is one of the important indexes to reflect the traffic state, and it is also one of the important contents in the research of the Intelligent Transportation system (its). The accurate prediction of road travel time is the important foundation of modern traffic guidance system and advanced traveler information system. Traffic data acquisition is the premise of road travel time prediction. The traditional traffic data acquisition methods are mainly based on induction coil, license plate recognition and GPS floating vehicle. Most of the previous studies on road travel time prediction mainly based on the traffic data collected by these three ways to build a model to predict the future road travel time. With the development of technology, especially the popularization of smart phone, the improvement of vehicle information system and the continuous progress of Bluetooth technology, the use of Bluetooth technology to collect road traffic data in real time has attracted much attention and been applied in some cities. The traffic data acquisition technology is mainly by detecting bluetooth signals in smart phones, bluetooth or Bluetooth headsets and other electronic devices on the road, and obtaining Bluetooth MAC addresses in these electronic devices. The MAC address matching is used to get the travel time and other traffic information in real time. This paper introduces in detail the principle of real-time traffic data acquisition based on Bluetooth technology, the selection of road travel time data based on Bluetooth technology and the estimation method of average travel time. The application of Bluetooth device in practice is investigated by questionnaire, and the proportion of Bluetooth signal detected in vehicle is estimated to be between 2.7% and 4.3%. In order to predict the future road travel time, this paper based on Kalman filter theory, support vector machine theory and particle swarm optimization algorithm. The short-term travel time prediction model based on Kalman filter and the road travel time prediction model based on support vector machine are established respectively. Based on the actual road travel time data collected by Bluetooth technology on the Humin Highway in Minhang District, Shanghai, the parameters of the prediction model, the training of the sample data set and the prediction solution are selected. The model is compared with the BP neural network model, and the results show that the proposed prediction model is effective.
【学位授予单位】：上海交通大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：U495

【相似文献】