A Predecessor Coding in an Evolutionary Algorithm for the Capacitated Minimum Spanning Tree Problem

This article presents an evolutionary algorithm (EA) for the capacitated minimum spanning tree problem occurring in telecommunication applications. The EA encodes a solution by a predecessor vector indicating for each node the preceding node at the path to the given central root node. Initialization, crossover, and mutation operators were specifically designed to provide strong locality and to enable an effective search in the space of feasible solutions only. Furthermore, local heuristics are applied to promote the inclusion of low-cost links. Empirical results on a set of standard test problems indicate that the EA performs better than two other heuristic techniques.