【作者】 姚俊杰；

【导师】 沈永增；

【作者基本信息】 浙江工业大学 ， 检测技术与自动化装置， 2007， 硕士

【摘要】 交通是国民经济发展的重要支柱。长期以来,交通问题已成为困扰城市发展的重要问题,尤其当驾驶者进入一个陌生的城市时,引导驾驶者走上正确的道路,及时准确地到达目的地就变得很重要。随着科技的飞速发展,计算机技术、网络技术和通讯技术已逐步渗入到交通领域,智能交通系统(ITS,Intelligent Transportation System)正是在这种情况下提出来的。交通地理信息系统在ITS中占有举足轻重的作用,已成为地理信息系统(GIS,Geographic Information System)的一个重要研究方向。车辆导航系统是一种典型的交通地理信息系统,具有极大的市场前景和发展潜力,其中的路径规划模块作为整个导航系统的核心模块之一,在嵌入式设备里运行,与在PC(Personal Computer)机上操作相比,要求路径规划时间更短和数据存储空间足够小,这就对路径规划算法设计在时间和空间上提出了新的要求。本文的研究主要涉及嵌入式开发(主要是软件开发)、测绘学中的地理信息系统原理、数据组织方法、数字地图可视化、最佳路径规划等技术。本文在分析电子地图的数据构成和模型的基础上,从MapInfo格式的桌面电子地图的MIF文件中提取出所需的电子地图数据。以图论作为理论基础来分析道路网络,在比较了图的几种经典存储结构后,结合C++标准模板库(STL,Standard Template Library)提出了一种高效的道路网络数据组织方式。并以出行距离作为路阻权值,在分析、比较几种经典的最短路径算法的基础上,结合交通路网特有的特征,对启发式A~*算法在数据存储、搜索策略等方面进行了改进。在程序实现上,采用C++语言,数据初始化部分和算法实现部分分别在两个类:CwayPoint类和CAStar类中实现,减小了它们之间的耦合性,真正体现了大型软件编程的模块化思想。实现了基于嵌入式电子地图导航的基本功能:电子地图的放大、缩小、漫游、兴趣点简单查询、路径寻优等。最后,在我们建立的嵌入式开发环境下,验证了实现的电子地图和算法的可行性。本文的研究以linux操作系统和Qt/Embedded作为开发平台,充分利用了免费的资源,与现有的大部分以WINCE操作系统和EVC(Embedded Visual C++)为平台的导航系统相比,大大节省了开发成本。更多还原

【Abstract】 The traffic is an important pillar of national economy development. For a long time, the traffic problem has already become an important problem which troubled urban development. Especially when driving into an unfamiliar city, guiding motorists on the right path and reaching their destinations becomes very important. Along with the rapid development of science and technology, computer technology, internet technology and communication technology have infiltrated into traffic field gradually. ITS(Intelligent transportation system) is proposed precisely in this context. Traffic geographic information system play a pivotal role in ITS, and has become an important research direction of GIS(Geographic Information System). Vehicle navigation system is a typical example of Traffic geographic information system, has great market prospect and development potentiality. The path planning module, as a core module among whole navigation system, compared with operating on Personal Computer, running in embedded devices requires shorter time and datas storage space small enough, that result in new demands for path planning algorithm’s design in time and space.The paper mainly involved embedded development (mainly software development), geographic information system of topography, datas organization method, digital map visualization, best path planning technology etc.On the basis of analyzing the data forming and model of electronic map, the paper extracted necessary datas from MIF files of MapInfo format electronic map, analyzed the road network based on graph theory. After comparing several kinds of classic storage structure of graph, combined STL(Standard Template Library) effectively, puted forward a datas organization method of road network. And used distance as road resistance weight, on the basis of analyzing and comparing several classical shortest path algorithm, combined specific characteristic of traffic network, improved heuristic A~* algorithm in data storage, search strategies etc. Adopted C++ language in program implementation aspect, data initialization and algorithm was implemented in two different class respectively: CwayPoint class and CAStar class, reduced coupling of them, reflected the modular idea of large software programming really. Came ture the basic functions of embedded navigation electronic map: zoom out, zoom in, roaming, inquiry of interested points, route planning etc. Finally, gave the expected experiment results under the embedded development environment that we set up.The research of this paper used linux and Qt/Embedded as developing platform, taked advantage of free resources fully. Comparing with most existing navigation system which use WINCE operating system and EVC(Embedded Visual C++) as platform, saved cost of development greatly.更多还原