车辆导航系统中地图匹配的研究
|
摘要
随着我国交通系统的迅速发展,新建道路和其他的交通设施越来越多,车辆导航系统作为交通运输系统的发展方向之一,在提高我国交通系统运行能力及改善交通状况的过程中扮演着越来越重要的角色。其中,导航系统的定位准确性是关键的,只有准确地定位车辆位置,才能向司机提供准确的行驶路线以及在地图上正确地显示车辆位置。
如何向司机提供实时、准确的车辆位置是车辆导航系统的重点和难点。作为一种定位技术,地图匹配以电子地图的高精度道路信息为匹配的数据基础,将GPS接收机接收到的车辆位置或行驶轨迹数据,与车载电子地图道路数据相比较,从而找到车辆所在的道路,并显示车辆的实时位置。
本文主要研究车辆导航系统中的地图匹配部分。首先,在了解车辆导航系统及其模块组成的基础上,对地图匹配的基本概念,发展现状以及基本原理进行了详细的论述。在深入分析影响地图匹配实时性、鲁棒性和匹配精度的因素后,对常用的投影法、概率统计法、相关性算法、基于网络拓扑以及基于曲线拟合的地图匹配算法进行分析,并总结各算法的优缺点。之后,介绍本文使用的基于D-S证据推理的地图匹配算法,完成了地图匹配三个过程的设计,即:确定误差区域、选择匹配道路以及修正定位结果。针对D-S证据推理使用的单一决策标准,引入了Pignistic概率决策,进行更有效的道路判断。同时,利用空间道路的拓扑关系来加快道路搜索的速度和提高道路选择的正确性。对于匹配过程中直转弯路段、平行路段、交叉路口路段以及立交桥路段等特殊路段的匹配情况,设计不同的方法来完成匹配,对出现的不同异常情况给出了不同的处理方法。在该系统中,使用ZHN600GPS导航仪进行跑车试验,并用VISUAL GPS记录下各个采样点的信息。对结果中匹配误差比较的大的转弯路段、平行路段以及立交桥路段的匹配,进行MATLAB仿真,并给出相应的仿真结果图。结果表明,该算法具有较高的匹配精度,对于复杂路口也有较好的适应性。
论文最后总结了在地图匹配研究中所做的工作,并展望了下一步需要改进的地方。
As the rapid development of Chinese transport system, new roads and other transport facilities are increasing. As a direction of the transport system development, vehicle navigation system is important for improving the traffic situation. The accuracy of the positioning system is crucial. Accurate routes and the vehicle location displayed on the map could be provided for drivers correctly only if the vehicle location is known accurately.
How to provide drivers with real-time, accurate location of vehicles has become the focus of vehicle navigation systems. As a positioning technology, map matching is based on the high-precision road information of an electronic map. It compares the data of vehicle location or track received by the GPS receiver and the data of road information in an electronic map, and finds the road vehicles are driven on and display the vehicle location.
The paper is concentrated on map matching in vehicle navigation system. First, on the basis of understanding the vehicle navigation and its modules, the concepts and basic principles of map matching are discussed in detail. After deep analysis of factors of the real-time, the robust and the accuracy of map matching, the common algorithms are studied, and the advantages and disadvantages of them are summed up. These algorithms include the projection, probability and statistics algorithm, the correlation algorithm, the algorithm based on the network topology and the algorithm based on the curve fitting. Then the algorithm of map matching based D-S evidence reasoning is introduced. In this paper, that algorithm is used to complete three courses of map matching, that is, determining the regional of error, choosing the matching section and amending the results of orientation. As for the single decision-making standard in the D-S evidence reasoning, the decision-making of Pignistic probability is introduced to improve the accuracy of road choosing. At the same time, the space topological of roads is introduced to speed up the road searching and improve the accuracy of road choosing. Different methods are designed to complete the matching of the turning sections, the parallel sections, the crossroad sections and the overpass sections. Different approaches are used for different circumstances. In this system, ZH600GPS are used to complete the testing and VISUAL GPS are used to record the information of all sampling points. The special sections with big error, such as the turning sections, the parallel sections and the overpass sections, are simulated in MATLAB 7.0 to verify the effectiveness and rationality of the algorithm. The results show that the algorithm in this paper has high matching accuracy and better adaptability in complex sections.
In the end, we summarize the research work of map matching in this paper, and describe the prospect to improve the map match in the future.
如何向司机提供实时、准确的车辆位置是车辆导航系统的重点和难点。作为一种定位技术,地图匹配以电子地图的高精度道路信息为匹配的数据基础,将GPS接收机接收到的车辆位置或行驶轨迹数据,与车载电子地图道路数据相比较,从而找到车辆所在的道路,并显示车辆的实时位置。
本文主要研究车辆导航系统中的地图匹配部分。首先,在了解车辆导航系统及其模块组成的基础上,对地图匹配的基本概念,发展现状以及基本原理进行了详细的论述。在深入分析影响地图匹配实时性、鲁棒性和匹配精度的因素后,对常用的投影法、概率统计法、相关性算法、基于网络拓扑以及基于曲线拟合的地图匹配算法进行分析,并总结各算法的优缺点。之后,介绍本文使用的基于D-S证据推理的地图匹配算法,完成了地图匹配三个过程的设计,即:确定误差区域、选择匹配道路以及修正定位结果。针对D-S证据推理使用的单一决策标准,引入了Pignistic概率决策,进行更有效的道路判断。同时,利用空间道路的拓扑关系来加快道路搜索的速度和提高道路选择的正确性。对于匹配过程中直转弯路段、平行路段、交叉路口路段以及立交桥路段等特殊路段的匹配情况,设计不同的方法来完成匹配,对出现的不同异常情况给出了不同的处理方法。在该系统中,使用ZHN600GPS导航仪进行跑车试验,并用VISUAL GPS记录下各个采样点的信息。对结果中匹配误差比较的大的转弯路段、平行路段以及立交桥路段的匹配,进行MATLAB仿真,并给出相应的仿真结果图。结果表明,该算法具有较高的匹配精度,对于复杂路口也有较好的适应性。
论文最后总结了在地图匹配研究中所做的工作,并展望了下一步需要改进的地方。
As the rapid development of Chinese transport system, new roads and other transport facilities are increasing. As a direction of the transport system development, vehicle navigation system is important for improving the traffic situation. The accuracy of the positioning system is crucial. Accurate routes and the vehicle location displayed on the map could be provided for drivers correctly only if the vehicle location is known accurately.
How to provide drivers with real-time, accurate location of vehicles has become the focus of vehicle navigation systems. As a positioning technology, map matching is based on the high-precision road information of an electronic map. It compares the data of vehicle location or track received by the GPS receiver and the data of road information in an electronic map, and finds the road vehicles are driven on and display the vehicle location.
The paper is concentrated on map matching in vehicle navigation system. First, on the basis of understanding the vehicle navigation and its modules, the concepts and basic principles of map matching are discussed in detail. After deep analysis of factors of the real-time, the robust and the accuracy of map matching, the common algorithms are studied, and the advantages and disadvantages of them are summed up. These algorithms include the projection, probability and statistics algorithm, the correlation algorithm, the algorithm based on the network topology and the algorithm based on the curve fitting. Then the algorithm of map matching based D-S evidence reasoning is introduced. In this paper, that algorithm is used to complete three courses of map matching, that is, determining the regional of error, choosing the matching section and amending the results of orientation. As for the single decision-making standard in the D-S evidence reasoning, the decision-making of Pignistic probability is introduced to improve the accuracy of road choosing. At the same time, the space topological of roads is introduced to speed up the road searching and improve the accuracy of road choosing. Different methods are designed to complete the matching of the turning sections, the parallel sections, the crossroad sections and the overpass sections. Different approaches are used for different circumstances. In this system, ZH600GPS are used to complete the testing and VISUAL GPS are used to record the information of all sampling points. The special sections with big error, such as the turning sections, the parallel sections and the overpass sections, are simulated in MATLAB 7.0 to verify the effectiveness and rationality of the algorithm. The results show that the algorithm in this paper has high matching accuracy and better adaptability in complex sections.
In the end, we summarize the research work of map matching in this paper, and describe the prospect to improve the map match in the future.
引文
[1]姚新春.多传感器集成定位技术在汽车导航中的应用.硕士学位论文.武汉大学.2004年2月.66-68
[2]http://www.cuauto.com.cn/Jsyj/News_Content.asp?Id=677
[3]孟炳林,高博.VNS中几个关键的GIS问题的研究.郑州信息工程大学学报.2001年12月.第2卷,第4期.66-68
[4]慈炳坤.车辆自主导航系统数字电子地图的研究.大连理工大学学报.2000年3月.7-11
[5]常菲.现代物流中车辆智能定位与导航系统中关键技术研究.上海交通大学学报.2004年1月.13-16
[6]张振辉.车辆导航系统中地图匹配算法与研究.硕士学位论文.郑州信息工程大学.2006年2月.11-15
[7]Duan Lingyu,Zhang Wangsheng.GPS Vehicle Navigationg System.Transactions of Nanjing University of Aeronautics & Astronautics.1998.15(2).Page:66-75
[8]JoshiR.A new approach to map matching in vehicle navigation systems.Intelligent Transportation System Conference Proceedings.2001.Oakland.USA.IEEE.2001
[9]张巨,景宁,刘雨,陈宏盛.GPS/GIS车辆路径定位方法研究.遥测遥控.1999年20卷第5期.16-17
[10]李锐明.基于预测的地图匹配方法.计算机工程与应用.2003第39卷第20期.110-113
[11]徐爱功,刘经南,王宏伟.车辆导航定位中地图匹配的MP模型.辽宁工程技术大学学报.2000年19卷第5期.20-24
[12]张小国.车载组合导航系统中的分级地图匹配算.中国惯性技术报.2000年8卷第3期.81-83
[13]柏钢.一种改进的地图匹配算法在车辆导航中的应用.中国惯性学报.1999年第7卷第1期.66-72
[14]张守信.GPS技术与应用.北京:国防工业出版社.1999
[15]毕军,付梦印,张宇河.基于D-S证据推理的车辆导航系统地图匹配算法.北京理工大学学报.2002年6月.21-24
[16]丁捷,杨殿阁,颜波.汽车导航用电子地图的高效路网模型.汽车工程.2003年第25卷第3期.122-130
[17]李云洁.汽车导航系统中地图匹配算法的研究与实现.硕士学位论文.吉林大学.2006年4月.32-37
[18]许志海.车辆导航系统中定位数据处理和地图匹配技术.测绘学院学报.2003年4月.331-334
[19]张其善,吴今培,杨东凯.智能车辆定位导航系统及应用.北京:科学出版社.2002年4月
[20]朱华统、黄继文.椭球大地计算.北京:八一出版社.1993年第一版
[21]LaurenceW,CarstensenJr.GPS and GIS.Enhanced Accuracy in MapMatching through Effective Filtering of Autonomous GPS Points.Cartography and Geographic Information Systems.1998.25(1).Page:51-62
[22]彭飞,柳重堪,张其善.基于模糊逻辑的GPS/DR组合导航系统地图匹配算法.遥测遥控.2001年1月.84-88
[23]张振辉,崔铁军,姚慧敏.车辆导航系统中地图匹配新算法.海洋测绘.2006年第26卷第2期.31-34
[24]周颖.基于曲线拟合的地图匹配算法.交通运输系统工程与信息.2004年5月.51-57
[25]陈佳瑜,肖桂荣.基于权重的地图匹配算法.计算机工程与应用.2005年11月.9-13
[26]孙世博.基于GPRS的GPS车辆定位系统及其证据推理地图匹配算法.哈尔滨工业大学学报.2006年6月.13-16
[27]常青,杨东凯,寇艳红,张其善.车辆导航定位方法及应用.北京:机械工业出版社.2005年5月
[28]Bullock JB,Krakiwsky EJ.Analysis of the Use of Digital Road Maps in Vehicle Navigation[C].In:ProcEEE Position Location and Navigation Symposium.1994.Page:494-501
[29]Sudano,JohnJ.Pignistic Probability Transforms for Mixes of Low-and-high-probability Events[C].The Fourth International Conference on Information Fusion Montreal PC.Canada.2001.Page:167-170
[2]http://www.cuauto.com.cn/Jsyj/News_Content.asp?Id=677
[3]孟炳林,高博.VNS中几个关键的GIS问题的研究.郑州信息工程大学学报.2001年12月.第2卷,第4期.66-68
[4]慈炳坤.车辆自主导航系统数字电子地图的研究.大连理工大学学报.2000年3月.7-11
[5]常菲.现代物流中车辆智能定位与导航系统中关键技术研究.上海交通大学学报.2004年1月.13-16
[6]张振辉.车辆导航系统中地图匹配算法与研究.硕士学位论文.郑州信息工程大学.2006年2月.11-15
[7]Duan Lingyu,Zhang Wangsheng.GPS Vehicle Navigationg System.Transactions of Nanjing University of Aeronautics & Astronautics.1998.15(2).Page:66-75
[8]JoshiR.A new approach to map matching in vehicle navigation systems.Intelligent Transportation System Conference Proceedings.2001.Oakland.USA.IEEE.2001
[9]张巨,景宁,刘雨,陈宏盛.GPS/GIS车辆路径定位方法研究.遥测遥控.1999年20卷第5期.16-17
[10]李锐明.基于预测的地图匹配方法.计算机工程与应用.2003第39卷第20期.110-113
[11]徐爱功,刘经南,王宏伟.车辆导航定位中地图匹配的MP模型.辽宁工程技术大学学报.2000年19卷第5期.20-24
[12]张小国.车载组合导航系统中的分级地图匹配算.中国惯性技术报.2000年8卷第3期.81-83
[13]柏钢.一种改进的地图匹配算法在车辆导航中的应用.中国惯性学报.1999年第7卷第1期.66-72
[14]张守信.GPS技术与应用.北京:国防工业出版社.1999
[15]毕军,付梦印,张宇河.基于D-S证据推理的车辆导航系统地图匹配算法.北京理工大学学报.2002年6月.21-24
[16]丁捷,杨殿阁,颜波.汽车导航用电子地图的高效路网模型.汽车工程.2003年第25卷第3期.122-130
[17]李云洁.汽车导航系统中地图匹配算法的研究与实现.硕士学位论文.吉林大学.2006年4月.32-37
[18]许志海.车辆导航系统中定位数据处理和地图匹配技术.测绘学院学报.2003年4月.331-334
[19]张其善,吴今培,杨东凯.智能车辆定位导航系统及应用.北京:科学出版社.2002年4月
[20]朱华统、黄继文.椭球大地计算.北京:八一出版社.1993年第一版
[21]LaurenceW,CarstensenJr.GPS and GIS.Enhanced Accuracy in MapMatching through Effective Filtering of Autonomous GPS Points.Cartography and Geographic Information Systems.1998.25(1).Page:51-62
[22]彭飞,柳重堪,张其善.基于模糊逻辑的GPS/DR组合导航系统地图匹配算法.遥测遥控.2001年1月.84-88
[23]张振辉,崔铁军,姚慧敏.车辆导航系统中地图匹配新算法.海洋测绘.2006年第26卷第2期.31-34
[24]周颖.基于曲线拟合的地图匹配算法.交通运输系统工程与信息.2004年5月.51-57
[25]陈佳瑜,肖桂荣.基于权重的地图匹配算法.计算机工程与应用.2005年11月.9-13
[26]孙世博.基于GPRS的GPS车辆定位系统及其证据推理地图匹配算法.哈尔滨工业大学学报.2006年6月.13-16
[27]常青,杨东凯,寇艳红,张其善.车辆导航定位方法及应用.北京:机械工业出版社.2005年5月
[28]Bullock JB,Krakiwsky EJ.Analysis of the Use of Digital Road Maps in Vehicle Navigation[C].In:ProcEEE Position Location and Navigation Symposium.1994.Page:494-501
[29]Sudano,JohnJ.Pignistic Probability Transforms for Mixes of Low-and-high-probability Events[C].The Fourth International Conference on Information Fusion Montreal PC.Canada.2001.Page:167-170