CAJ | 학술논문

针对量子粒子群优化算法早熟收敛的问题，提出一种双心扰动的变异机制。对粒子的势能中心和粒子群的重心进行自适应柯西变异，发挥两者在进化后期的协同引导能力，以提高进化后期粒子群对新空间的开拓能力。对4个典型测试函数进行仿真实验，结果表明，对于单峰函数优化，双心扰动变异机制的优化效果比只采用势能中心、重心和全局最好位置变异的优化效果提高36.42%以上；对于多峰函数优化，其优化效果提高32.84%以上。
침대양자입자군우화산법조숙수렴적문제，제출일충쌍심우동적변이궤제。대입자적세능중심화입자군적중심진행자괄응가서변이，발휘량자재진화후기적협동인도능력，이제고진화후기입자군대신공간적개탁능력。대4개전형측시함수진행방진실험，결과표명，대우단봉함수우화，쌍심우동변이궤제적우화효과비지채용세능중심、중심화전국최호위치변이적우화효과제고36.42%이상；대우다봉함수우화，기우화효과제고32.84%이상。
Aiming at the problem of the premature convergence of Quantum Particle Swarm Optimization(QPSO) algorithm. This paper introduces a double core disturbance mutation mechanism. It uses adaptive Cauchy mutation to mutate the potential energy of the particle center and the center of gravity of the particle swarm and make full use of the guiding ability of the two centers in the late part of evolution. It adopts four typical functions to conduct simulation experiment, results show that double core disturbances mutation mechanism optimization is better than the strategy of only potential energy center, the center of gravity or global optimal mutation at least 36.42%, and the optimization results improve at least 32.84%for multimodal function optimization.