基于手机位置信息的地图匹配算法研究

发布时间：2018-05-31 09:04

本文选题：智能交通 + GPS定位　；参考：《大连理工大学》2015年硕士论文

【摘要】：近年来,我国城市交通拥堵问题越来越严重,虽然客观上存在基础建设发展滞后、路网规划不合理等因素,但主观上也存在调度、使用不合理的问题,利用智能交通系统解决城市交通拥堵问题已越来越被人们关注。通过智能系统管理城市交通,重点是及时准确地获取车辆的定位信息。获取车辆位置信息常用的方法有二种,一种是传统方法,包括安装检测器和定点上报,另一种是现代方法,包括基于GPS路况信息监测、众包、统计预测、蓝牙定位等方法。这些方法在设备安装和维护上投入相对较大,耗费较多的人力物力和财力,不易于推广普及,这对采集信息的准确性和全面性有较大影响。相比以上方法,利用手机获取位置数据在采样点数量、覆盖率、开销等方面具有优势。本文在分析当前手机定位方法优缺点的基础上,重点对利用手机基站获取车辆位置信息的方法进行研究,并用实际采集的位置信息进行了检验。利用手机获取位置信息通常有两种方法,一种是利用手机GPS获取位置信息,一种是利用手机基站获取位置信息。在对比两种方法时发现,利用手机GPS获取位置信息能耗太大,严重缩短手机的待机时间,非常不利于推广普及,也就很难发挥在采样点数量和覆盖率上的优势；利用手机基站获取位置信息虽然能耗低,对手机的待机时间基本没有影响,但其定位信息的精度较差。而定位的准确性是智能交通系统的重要因素,为了克服定位误差,提供可靠、准确的定位数据,需要通过使用地图匹配算法将车辆定位点精确地匹配到交通数字地图的道路网中。现有的地图匹配算法大都在GPS定位上得到了很好的应用,但能否在手机基站定位上应用尚不可知。为此,我们引入一种较为适应手机基站定位误差特性的隐马尔可夫地图匹配算法来提高手机基站定位的精度,算法将定位点匹配到候选匹配道路上的概率称作初始匹配概率,将候选匹配道路发生转换的概率称作道路转移概率,算法使用动态编程来快速找到初始匹配概率和道路转移概率乘积最大的路径。考虑到车辆在实际行驶过程中不可能频繁地转换道路,我们在算法计算定位点道路转移概率时引入了约束限制进行改进,从而提高了算法性能。为了检验算法在真实数据中的运行情况以及能否应用于手机定位,在用手机采集大量定位数据的基础上,对算法分别用采集的GPS定位数据和手机基站定位数据进行了检验。结果显示,该算法对手机GPS定位点处理较好,但对手机基站定位点进行处理时,输出结果不稳定。实验结果表明,可以通过提高采样频率来提高手机基站定位精度。
[Abstract]:In recent years, the problem of urban traffic congestion in China has become more and more serious. Although there are some factors, such as lagged infrastructure development and unreasonable road network planning, there are also subjective problems of scheduling and unreasonable use. People pay more and more attention to using intelligent transportation system to solve the problem of urban traffic congestion. Through the intelligent system to manage the urban traffic, the emphasis is to get the location information of the vehicle in time and accurately. There are two common methods for obtaining vehicle position information, one is traditional method, which includes installing detector and fixed point reporting, the other is modern method, including monitoring of road condition based on GPS, crowdsourcing, statistical prediction, Bluetooth location and so on. These methods are relatively large in installation and maintenance of equipment, expend more manpower, material and financial resources, and are not easy to be popularized, which has a great impact on the accuracy and comprehensiveness of collecting information. Compared with the above methods, the use of mobile phone to obtain location data in the number of sampling points, coverage, cost and other advantages. Based on the analysis of the advantages and disadvantages of the current mobile phone positioning methods, this paper focuses on the research on the method of obtaining the vehicle location information by using the mobile phone base station, and tests it with the actual collected location information. There are usually two methods to obtain location information by using mobile phone, one is to obtain location information by mobile phone GPS, and the other is to obtain location information by mobile phone base station. When comparing the two methods, it is found that the use of mobile phone GPS to obtain location information consumes too much energy, seriously shortens the standby time of the mobile phone, which is not conducive to popularization, and it is difficult to give full play to the advantages in the number and coverage of sampling points. Although the use of mobile phone base station to obtain location information has low energy consumption and has no effect on the standby time of mobile phone, the accuracy of location information is poor. The accuracy of location is an important factor in intelligent transportation system. In order to overcome the positioning errors and provide reliable and accurate positioning data, The vehicle location points need to be accurately matched to the road network of the traffic digital map by using the map matching algorithm. Most of the existing map matching algorithms have been well applied in GPS location, but it is not known whether they can be used in mobile phone base station location. For this reason, we introduce a hidden Markov map matching algorithm, which adapts to the location error of mobile base station, to improve the accuracy of mobile phone base station location. The probability that the location points are matched to the candidate matching path is called the initial matching probability. The transition probability of candidate matching road is called road transition probability. The algorithm uses dynamic programming to quickly find the path with the largest product of initial matching probability and road transition probability. In view of the fact that it is not possible for vehicles to switch roads frequently in the actual driving process, we introduce constraints to improve the algorithm performance when we calculate the road transition probability of positioning points. In order to verify the operation of the algorithm in the real data and whether it can be applied to mobile phone positioning, the algorithm is tested with the collected GPS location data and the mobile phone base station location data on the basis of a large number of location data collected by the mobile phone. The results show that the algorithm can deal with the mobile phone GPS location point well, but the output result is unstable when the mobile phone base station positioning point is processed. The experimental results show that the location accuracy of mobile phone base station can be improved by increasing the sampling frequency.
【学位授予单位】：大连理工大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：U495;TP301.6

【共引文献】