標題: A hybrid shortest path algorithm for navigation system
作者: Cho, Hsun-Jung
Lan, Chien-Lun
運輸與物流管理系 註:原交通所+運管所
Department of Transportation and Logistics Management
關鍵字: shortest path algorithm;heuristics;hierarchical network
公開日期: 2007
摘要: Combined with Geographic Information System (GIS) and Global Positioning System (GPS), the vehicle navigation system had become a quite popular product in daily life. A key component of the navigation system is the Shortest Path Algorithm. Navigation in real world must face a network consists of tens of thousands nodes and links, and even more. Under the limited computation capability of vehicle navigation equipment, it is difficult to satisfy the real-time response requirement that user expected. Hence, this study focused on shortest path algorithm that enhances the computation speed with less memory requirement. Several well-known algorithms such as Dijkstra, A* and hierarchical concepts were integrated to build hybrid algorithms that reduce searching space and improve searching speed. Numerical examples were conducted on Taiwan highway network that consists of more than four hundred thousands of links and nearly three hundred thousands of nodes. This real network was divided into two connected sub-networks (layers). The upper layer is constructed by freeways and expressways- the lower layer is constructed by local networks. Test origin-destination pairs were chosen randomly and divided into three distance categories; short, medium and long distances. The evaluation of outcome is judged by actual length and travel time. The numerical example reveals that the hybrid algorithm proposed by this research might be tens of thousands times faster than traditional Dijkstra algorithm; the memory requirement of the hybrid algorithm is also much smaller than the tradition algorithm. This outcome shows that this proposed algorithm would have an advantage over vehicle navigation system.
URI: http://hdl.handle.net/11536/8612
ISBN: 978-0-7354-0476-2
ISSN: 0094-243X
期刊: COMPUTATION IN MODERN SCIENCE AND ENGINEERING VOL 2, PTS A AND B
Volume: 2
起始頁: 1024
結束頁: 1027
Appears in Collections:Conferences Paper