Solving the hierarchical Chinese postman problem as a rural postman problem

Abstract

In the undirected hierarchical Chinese postman problem (HCPP), the edges of a graph are partitioned into clusters and must be serviced while respecting a hierarchy, or precedence relation, between clusters. Two objectives are considered: a hierarchical objective and a makespan objective. In this article, a transformation of the HCPP into an equivalent rural postman problem (RPP) is presented. The HCPP is solved optimally, for both objectives, by applying an exact branch-and-cut RPP algorithm to the transformed problem. Two heuristics based on the RPP algorithm are also described and assessed computationally.