A Dual Learning Strategy-based Particle Swarm Optimization with Adaptive Velocity Control
A Dual Learning Strategy-based Particle Swarm Optimization with Adaptive Velocity Control
复制标题
DOI：
10.1109/icnsc55942.2022.10004092
复制
发表时间：
2022-12
期刊：
影响因子：
--
通讯作者：
X. Yang;Zonghui Cai;Shangce Gao
中科院分区：
文献类型：
--
作者：
X. Yang;Zonghui Cai;Shangce Gao
粒子群优化算法由于其简单、有效的特点，在近几十年来受到了越来越多的关注。近年来，已经提出了许多变体的粒子群优化算法来解决各种优化问题。粒子之间的学习策略显着影响它们的搜索行为和性能。虽然各种新颖的学习策略已经被引入到粒子群优化算法中，但大多数都只使用一种机制。在本文中，我们提出了一种基于双学习策略的粒子群优化与自适应速度控制。该算法分别得益于综合学习和遗传学习，具有良好的探索性和开发性。此外，这两种能力很好地平衡，通过自适应参数和速度控制。在30个常用的基准函数上对该算法进行了测试，实验结果表明，该算法在有效性和鲁棒性方面优于其他基于学习策略的粒子群优化算法.
Particle swarm optimization has been paid more attention in recent decades, since its simplicity and effectiveness. Numerous variants of particle swarm optimization have been proposed for solving various optimization problems in recent years. Learning strategy between particles significantly affects their search behaviour and performance. Although various novel learning strategies have been introduced into particle swarm optimization, most of them are only use one type of mechanisms. In this paper, we propose a dual learning strategy-based particle swarm optimization with adaptive velocity control. The proposed algorithm has good exploration and exploitation which is benefited from comprehensive learning and genetic learning, respectively. Furthermore, these two abilities are well balanced according to adaptive parameter and velocity control. The proposed algorithm is tested on 30 frequently used benchmark functions, and the experimental results show that it outperforms other learning strategy-based particle swarm optimization algorithms in terms of effectiveness and robustness.