MH-MOEA: A New Multi-Objective Evolutionary Algorithm Based on the Maximin Fitness Function and the Hypervolume Indicator

In this paper, we propose an approach that combines a modified version of the maximin fitness function and the hypervolume indicator for selecting individuals into a Multi-Objective Evolutionary Algorithm (MOEA). Our proposed selection mechanism is incorporated into a MOEA which adopts the crossover and mutation operators of the Nondominated Sorting Genetic Algorithm-II (NSGA-II), giving rise to the so-called “Maximin-Hypervolume Multi-Objective Evolutionary Algorithm (MH-MOEA)”. Our proposed MH-MOEA is validated using standard test problems taken from the specialized literature, using from three to six objectives. Our results are compared with respect to those produced by MC-MOEA (which is based on the maximin fitness function and a clustering technique), MOEA/D using Penalty Boundary Intersection (PBI), which is based on decomposition and iSMS-EMOA (which is based on the hypervolume indicator). Our preliminary results indicate that our proposed MH-MOEA is a good alternative to solve multi-objective optimization problems having both low dimensionality and high dimensionality in objective function space.

[1]  Lakhmi C. Jain,et al.  Evolutionary Multiobjective Optimization , 2005, Evolutionary Multiobjective Optimization.

[2]  Xiaodong Li,et al.  Better Spread and Convergence: Particle Swarm Multiobjective Optimization Using the Maximin Fitness Function , 2004, GECCO.

[3]  Xiaoxia Huang,et al.  Optimizing expressway maintenance planning by coupling ant algorithm and geography information system transportation in Hubei province, China , 2011, 2011 IEEE International Geoscience and Remote Sensing Symposium.

[4]  Adriana Menchaca-Mendez,et al.  A new selection mechanism based on hypervolume and its locality property , 2013, 2013 IEEE Congress on Evolutionary Computation.

[5]  Anne Auger,et al.  Theory of the hypervolume indicator: optimal μ-distributions and the choice of the reference point , 2009, FOGA '09.

[6]  Adriana Menchaca-Mendez,et al.  Selection Operators Based on Maximin Fitness Function for Multi-Objective Evolutionary Algorithms , 2013, EMO.

[7]  Qingfu Zhang,et al.  MOEA/D: A Multiobjective Evolutionary Algorithm Based on Decomposition , 2007, IEEE Transactions on Evolutionary Computation.

[8]  Marco Laumanns,et al.  Scalable Test Problems for Evolutionary Multiobjective Optimization , 2005, Evolutionary Multiobjective Optimization.

[9]  John E. Dennis,et al.  Normal-Boundary Intersection: A New Method for Generating the Pareto Surface in Nonlinear Multicriteria Optimization Problems , 1998, SIAM J. Optim..

[10]  Nicola Beume,et al.  An EMO Algorithm Using the Hypervolume Measure as Selection Criterion , 2005, EMO.

[11]  Xiaodong Li,et al.  On performance metrics and particle swarm methods for dynamic multiobjective optimization problems , 2007, 2007 IEEE Congress on Evolutionary Computation.

[12]  R. J. Balling,et al.  The maximin fitness function for multi-objective evolutionary computation: application to city planning , 2001 .

[13]  Marco Laumanns,et al.  Performance assessment of multiobjective optimizers: an analysis and review , 2003, IEEE Trans. Evol. Comput..

[14]  Adriana Menchaca-Mendez,et al.  Solving multi-objective optimization problems using differential evolution and a maximin selection criterion , 2012, 2012 IEEE Congress on Evolutionary Computation.

[15]  Anne Auger,et al.  Theoretically Investigating Optimal µ-Distributions for the Hypervolume Indicator: First Results for Three Objectives , 2010, PPSN.

[16]  Lothar Thiele,et al.  Multiobjective Optimization Using Evolutionary Algorithms - A Comparative Case Study , 1998, PPSN.

[17]  Richard Balling,et al.  The Maximin Fitness Function; Multi-objective City and Regional Planning , 2003, EMO.

[18]  R. Lyndon While,et al.  A review of multiobjective test problems and a scalable test problem toolkit , 2006, IEEE Transactions on Evolutionary Computation.