Cognitive studies show that human being's path planning is generally conducted through opportunistic planning processes in which optimal paths are iteratively and dynamically constructed according to the changes of the traffic conditions, rather than once determined before departure. The optimal path finding system in transportation should reflect the cognitive process of human being's planning to achieve efficient and useful results in real-world situation. In this research, for the implementation of the planning concept in ITS, a blackboard model is employed to reflect human being's planning processes into automatic path finding. The blackboard architecture typically consists of knowledge sources, controller, and blackboard. In this research, a prototype system, SHARP (System for Highly Adaptive Route Planning) is developed to implement the blackboard architecture in which knowledge sources are optimal path finding module (SPFM), travel time forecasting module (TTFM), and incident detecting module (IDM). In a series of experimental tests, reduction of total travel time was achieved: the travel time of optimal path provided by traditional approach is 28.142, while the travel time of SHARP is 17.884. Also, SHARP can guarantee the optimality of path by updating the optimal path during driving according to the changes of traffic situations. Thus, SHARP is a system that properly provides optimal paths in real world situations. (A*)
Abstract