面向路网环境的智能路径搜索与导航技术研究

路径规划与导航是智能交通中最具代表性的一类应用，深入研究导航技术既能提高交通工具的使用效率，也可以减少道路拥挤，降低环境污染，提高城市居民的出行舒适度。然而，现有导航应用需要用户机械地输入起点与终点，然后简单描述起点与终点之间的最短或者最快路径。因此，其无法满足用户多样化的复杂搜索意图。本项目旨在提出一套完整的路径规划与导航的计算框架。利用数据库索引、路网与轨迹数据融合、自然语言处理和深度学习等相关技术，实现导航技术智能化。本项目将重点研究：1）基于模糊线索的自动路径搜索；2）面向行为轨迹示例的路径推荐；3）路网环境下的交互式路径导航。本项目的研究对推动智能化导航技术的进一步发展具有重要的科学意义，同时对大数据时代下的互联网各类应用的发展提供了科学技术层面的支撑。

Route planning and navigation is an important application in intelligent transportation. In-depth study of navigation technology cannot only improve the efficiency of vehicle use, but also reduce the road congestion and environmental pollution, and improve the comforts of urban residents. However, existing navigation applications require the user to manually input the source and destination, and then simply describe the shortest or fastest path between the source and destination. Therefore, it cannot satisfy the user’s diversified and complex search intent. This project aims to propose a complete computational framework for route planning and navigation. By using techniques such as database index, road network and trajectory data fusion, natural language processing and deep learning, the goal is to accomplish the intelligent navigation technology. The project will focus on: 1) automatic route search based on fuzzy clues; 2) route recommendation based on activity trajectory exemplar; 3) interactive path navigation in road network. This project has important scientific significance for promoting the further development of intelligent navigation technology. At the same time, it provides scientific and technical support for the development of various Internet applications under the era of big data.