A fast interpolation-based multi-objective evolutionary algorithm for large-scale multi-objective optimization problems

Abstract

Abstract Evaluating large-scale multi-objective problems is usually time-consuming due to a large number of decision variables. However, most of the existing algorithms for large-scale multi-objective optimization require a large number of problem evaluations to obtain acceptable results, which makes the optimization very inefficient. In this paper, a fast interpolation-based multi-objective evolutionary algorithm is proposed for solving large-scale multi-objective optimization problems with high convergence speed and accuracy. In the proposed algorithm, the decision variables are generated based on the information of a small number of variables by the interpolation function. With this approach, only a small number of variables need to be optimized in the proposed algorithm, and the search space can be reduced greatly to improve the convergence speed, and to make it possible to obtain satisfactory results with a relatively small computation cost. The experimental results verified that our proposed algorithm outperforms other compared algorithms in terms of convergence speed and convergence accuracy on 108 test instances with up to 1000 decision variables. Additionally, a parametric study is provided to investigate the best parameter setting for the proposed algorithm.