SGER: Map-Matching and Reactive Routing Algorithms for Traffic Estimation and Prediction Systems

地图匹配和路由算法是流量估计和预测系统 (TREPS) 的重要技术组件。 GPS 跟踪数据用于评估交通状况，但噪音和错误需要复杂的地图匹配算法来识别道路网络中的相应路线。 当前的地图匹配算法对于实时跟踪来说太慢且不准确，并且没有经过质量测试。 当前的路由算法基于存储每个路段的静态旅行时间权重的数据库，尽管交通状况的持续和快速变化需要结合动态变化的权重。 该项目 (i) 采用非常规方法来设计更快、更准确的地图匹配算法，包括评估其质量的新颖想法，以及 (ii) 启动反应式路由算法的研究，使其解决方案适应动态变化的权重数据库，包括支持性新颖数据库技术的研究。这项研究涉及以可证明正确的方式本地化高质量的全球地图匹配算法，并引入基于人工 GPS 数据的质量评估。 将启动反应式路由算法的研究，这需要新颖的数据仓库技术来联合管理历史和当前时空数据。这项研究将极大地增强 TREPS 以及数据库和算法研究社区的知识库。个人路线和导航解决方案领域正在迅速崛起，因为每部手机都有可能变成导航辅助设备。该项目的成果在这个​​蓬勃发展的地区以及交通评估和预测等领域具有很高的利用潜力，这些任务由市政当局和部委等交通管理机构执行。 这项研究涉及希腊和美国之间的探索性国际合作。这项资助由 OISE、欧洲和欧亚大陆以及 CCF/TF Cluster 共同资助。

Map-matching and routing algorithms are essential technical components for Traffic Estimation and Prediction Systems (TREPS). GPS tracking data is used to assess traffic condition, however noise and errors require sophisticated map-matching algorithms to identify corresponding routes in the road network. Current map-matching algorithms are too slow and inaccurate for real-time tracking, and are not quality-tested. Current routing algorithms are based on databases which store static travel-time weights per road segment, although continuous and rapid changes in traffic conditions demand the incorporation of dynamically changing weights. This project (i) follows unconventional approaches for the design of faster and more accurate map-matching algorithms, including novel ideas to evaluate their quality, and (ii) initiates research of reactive routing algorithms that adapt their solution to a dynamically changing weight database, including research of supportive novel database techniques.This research involves localizing high-quality global map-matching algorithms in a provably correct manner, and introduces quality evaluation based on artificial GPS data. Research into reactive routing algorithms will be initiated, which requires novel data warehousing techniques for the joint management of historic and current spatiotemporal data. This research will greatly enhance the knowledge base for TREPS and for both database and algorithms research communities. The area of personal routing and navigation solutions is quickly on the rise since every mobile phone can potentially be turned into a navigational aid. Results of this project have a high potential for utilization in this booming area and in areas such as traffic assessment and prediction, tasks that are performed by traffic management authorities such as municipalities and ministries. This research involves an exploratory international collaboration between Greece and the US.This grant is jointly funded by OISE, Europe and Eurasia, and CCF/TF Cluster.