Multi Mixed Population Evolutionary Algorithm applied to the Multiobjective Degree Constrained Minimum Spanning Tree Problem
Genetic Algorithms, Hybrid Genetic Algorithms, Multiobjective Programming, Degree Constrained Minimum Spanning Tree.
In recent years, the Multiobjective Degree Constrained Minimum Spanning Tree Problem, has been gaining the attention of researchers in Combinatorial Optimization area, due to its wide application in practical real-world problems, specially network modeling related. It is considered a NP-hard problem, even in its mono-objective version for a degree of at least . This research proposes to solve the problem employing a new evolutionary algorithm called MMPEA. This approach combines the particular characteristics of multiple hybrid populations in order to diversify the phenotype of individuals, to improve the survey of the search space. For generation of the three mixed populations used in this first version of the algorithm, the PAES, M-PAES and AESSEA algorithms were used, which were also used in comparison tests with the MMPEA. Due to the multi-objective nature of the problem, the results for these early experiments are presented through hypervolume and binary epsilon indicators.