基于分解的多目标量子差分进化算法

摘要

Multi-objective evolutionary algorithm based on decomposition (MOEA /D)is featured by high convergence rate and good distribution.However,its performance in non-convex functions is not good enough.In view of the excellent properties of quantum evolutionary algorithm in multi-peak functions,we combined MOEA /D with QEA and proposed the decomposition-based quantum differential multi-objective evolutionary algorithm (QD-MOEA /D).The quantum chromosome of QD-MOEA /D adopts real number in encoding,this saves memory space and accelerates operation speed.In order to speed up convergence speed and improve detection ability of algorithm,the quantum chromosome adopts differential evolution,and its mutation way is the quantum non-gate.Results of experiments on several standard test functions showed that the algorithm improved the convergence and the distribution of MOEA /D in non-convex functions.%基于分解的多目标进化算法 MOEA ／D（Multi-objective Evolutionary Algorithm Based on Decomposition）具有收敛速度快、分布性好等特点，但其在非凸函数上的性能有待提高。鉴于量子进化算法在多峰值函数上的优良性能，将 MOEA ／D 与量子进化算法相结合，提出基于分解的多目标量子差分进化算法 QD-MOEA ／D（Quantum Differential Multi-objective Evolutionary Algorithm Based on Decomposition）。QD-MOEA ／D 的量子染色体采用实数编码，节省存储空间，加快运算速度。为了加快算法收敛速度并提高算法探测能力，量子染色体采取差分进化，其变异方式为量子非门。在多个标准测试函数的实验结果表明，该算法改进了 MOEA ／D 在非凸函数上的收敛性和分布性。