A New Route Optimization Algorithm for Rapid Decision Support

Abstract

We describe a new heuristic search algorithm, Interruptible A (IA), that we have implemented in a real-world decision aid for use of public transit. IA is appropriate for shortest path problems where there is value to a suboptimal path returned quickly. We offer an example in which IA returns an optimal path in single iteration, and another where the algorithm finds a suboptimal path quickly before converging to the optimal path. Two admissibility condition are presented, along with empirical results indicating that IA is effective in both admissible and inadmissible cases.