Computational study of an improved shortest path algorithm

Shortest and/or longest path analysis is a major analytical component of quantitative models used by transportation planners. The importance of shortest path analysis in all phases of transportation planning has given rise to intensive research and software development over the last three decades for solving shortest path problems. This paper presents a new hybrid solution algorithm called THRESH, which integrates the features of label setting and label correcting algorithms, yet appears to have performance characteristics that transcend both. Preliminary computational results indicate that it is substantially more efficient than the methods determined to be best by previous studies.