基于多点非均匀变异的多目标极值优化算法研究

Research on Multi-objective Extremal Optimization Algorithm with Multi-non-uniform Mutation

摘要:

多目标进化算法因其在解决含有多个矛盾目标函数的多目标优化问题中的强大处理能力，正受到越来越多的关注与研究。极值优化作为一种新型的进化算法，已在各种离散优化、连续优化测试函数以及工程优化问题中得到了较为成功的应用，但有关多目标EO算法的研究却十分有限。本文将采用Pareto优化的基本原理引入到极值优化算法中，提出一种求解连续多目标优化问题的基于多点非均匀变异的多目标极值优化算法。通过对六个国际公认的连续多目标优化测试函数的仿真实验结果表明：本文提出算法相比NSGA-II、 PAES、SPEA和SPEA2等经典多目标优化算法在收敛性和分布性方面均具有优势。

Abstract:

Multi-objective evolutionary algorithms have attracted more and more attentions because their powerful solving ability of dealing with several conflict objective functions for a multi-objective optimization problem. As a novel evolutionary algorithm, extremal optimization has been applied successfully to a variety of discrete, continuous optimization test functions and engineering optimization problems, but there is limited research works concerning multi-objective extremal optimization. This paper introduces the basic ideas of Pareto optimization to extremal optimization, and proposes a multi-objective extremal optimization algorithm with multi-non-uniform mutation for solving the continuous multi-objective optimization problems. The simulation experiments on six well- known continuous multi-objective optimization test functions have demonstrated that the proposed algorithm in this paper is superior to other traditional multi-objective optimization algorithms such as NSGA-II, PAES, SPEA and SPEA2 in terms of convergence and diversity.