An Elite-Guided Evolutionary Algorithm for Large-Scale Multi-Objective Optimization

When handling large-scale multi-objective optimization problems, a good diversity maintenance can effectively avoid the population trapping into premature convergence and enhance the utilization of the decision space. However, the recombination operator in the existing multi-objective optimization algorithm is difficult to find the local optimal solution and the global optimal solution in the huge decision space due to the low search efficiency. To get the best trade-off between exploration and exploitation during the evolutionary search process, this paper proposes an elite-guided evolutionary algorithm for large-scale multi-objective optimization. The proposed algorithm adopts a recombination operator with a novel search strategy that explicitly utilizes local similarity neighborhood property between the population in the decision space and the objective space to guide the individuals to generate a diversity approximation of the Pareto front, which can highly promote the search efficiency. The experimental results on a variety of general large-scale benchmark problems demonstrate the competitiveness and effectiveness of the developed algorithm over several state-of-the-art multi-objective evolutionary algorithms.