摘要
为解决反向粒子群优化算法存在易陷入局部最优、收敛速度过慢的问题,文章提出了一种模拟退火自适应反向无惯性粒子群优化(simulated annealing adaptive opposition-based non-inertial particle swarm optimization,SAONPSO)算法。SAONPSO算法在一般性反向学习方法的基础上,根据模拟退火原理,对每一代粒子群自适应地给出反向策略执行概率,并结合精英差分变异策略对当前最优粒子进行扰动,从而帮助粒子避免陷入局部最优解;为加快粒子收敛速度,SAONPSO算法引入了一种无惯性速度更新公式来引导粒子飞行方向,该公式在对环境信息充分获取的基础上以较少的参数设置获得了较快的收敛速度。上述2种机制的结合应用有效地克服了反向粒子群算法中探索与开发的矛盾;实验结果表明,与其他反向粒子群优化算法相比,SAONPSO算法在解的求解精度与收敛速度上均较优。
To solve the problem of falling into local optimization and slow convergence speed in opposition-based particle swarm optimization,this paper proposes a simulated annealing adaptive opposition-based non-inertial particle swarm optimization(SAONPSO).The new algorithm presents opposition-based learning strategy execution probability according to the principle of simulated annealing(SA)on the basis of the generalized opposition-based learning(GOBL)strategy for each evolution generation,and combines an elite differential evolutionary mutation(EDEM)strategy to perturb the current global optimal particle,thus helping the particles avoid falling into local optimal position.In order to accelerate convergence speed,SAONPSO introduces a non-inertial velocity update formula(NIV)to lead flight direction of particles,which achieves faster convergence speed with fewer parameter settings on the basis of sufficient acquisition of environmental information.The combination of the above two strategies effectively balances the contradiction between exploration and exploitation.Experimental results show SAONPSO algorithm outperforms in calculation accuracy and convergence speed compared with the state-of-the-art opposition-based particle swarm optimizations.
作者
曹文梁
康岚兰
王石
CAO Wenliang;KANG Lanlan;WANG Shi(Department of Computer Engineering,Dongguan Polytechnic,Dongguan 523808,China;College of Applied Science,Jiangxi University of Science and Technology,Ganzhou 341000,China)
出处
《合肥工业大学学报（自然科学版）》
CAS
北大核心
2020年第12期1614-1619,共6页
Journal of Hefei University of Technology：Natural Science
基金
江西省科技厅自然科学基金面上资助项目(20202BABL202032)
广东省普通高校特色创新(自然科学)资助项目(2019GKTSCX142,2017GKTSCX101)
东莞职业技术学院技艺能手资金资料项目(Y17040323)。
关键词
一般性反向学习
粒子群优化
模拟退火
精英差分变异
无惯性速度更新公式
generalized opposition-based learning(GOBL)
particle swarm optimization(PSO)
simulated annealing(SA)
elite differential evolutionary mutation(EDEM)
non-inertial velocity update formula(NIV)
