求解动态优化问题的多种群骨干粒子群算法

发布时间：2018-05-01 05:04

本文选题：动态优化问题 + 骨干粒子群算法　；参考：《计算机工程与应用》2017年19期

【摘要】：针对动态优化问题(Dynamic Optimization Problem,DOP)中所面临的过时记忆和多样性丧失的挑战,提出了一种改进的多种群骨干粒子群优化算法(Multi-swarms Bare Bones Particle Swarm Optimization,MBBPSO)。通过设置环境勘探粒子及时检测环境的变化,避免了错误信息误导种群的进化方向;环境改变后,利用上一个环境搜索的信息初始化新的种群,提高MBBPSO快速追踪到当前环境的优秀解的能力;当种群陷入停滞时,采用新的进化方程以加强粒子的活性和多种群策略维持群体的多样性。仿真实验表明,MBBPSO在解决动态环境问题中具有较强的竞争力。
[Abstract]:Aiming at the challenge of out-of-date memory and loss of diversity in dynamic Optimization problem, an improved multi-swarms Bare Bones Particle Swarm optimization algorithm is proposed. By setting up environmental exploration particles to detect the changes in the environment in time, the wrong information is avoided to mislead the evolution direction of the population; after the environment changes, the new population is initialized with the information from the previous environmental search. Improve the ability of MBBPSO to track the excellent solutions to the current environment quickly; when the population is stagnant, the new evolution equation is adopted to enhance the activity of particles and maintain the diversity of the population. Simulation results show that MBBPSO is competitive in solving dynamic environmental problems.
【作者单位】：安徽工业大学计算机科学与技术学院;
【基金】：国家自然科学基金(No.61300059,No.61502010)
【分类号】：TP18

【相似文献】