基于天牛须搜索自适应的樽海鞘算法-《计算机技术与发展》

文章信息/Info

Author(s):: YANG Xing; GUO Ming-hao; FANG Xia; ZHU Zhong-ming; JIANG Mei-qi; School of Information Science and Technology,Chengdu University of Technology,Chengdu 610059,China

Keywords:: salp swarm algorithm; beetle antennae search; Lorentz function; adaptive; global optimal

摘要:: 针对樽海鞘算法在优化求解问题时收敛速度慢和局部优化能力差等缺点, 引入不同的优化策略对其进行改进,提出一种基于天牛须搜索自适应的樽海鞘算法。改进的樽海鞘算法在领导者位置更新中引入天牛须搜索机制和洛伦兹函数替代基本樽海鞘算法领导者位置更新公式中的随机值, 提高了算法的局部优化能力;在追随者位置更新中引入自适应惯性权重, 调节自身位置和上一代位置对追随者位置更新的影响,在全局搜索和局部搜索之间提供更好的平衡。将改进的樽海鞘算法通过对 8 个不同类型的基准测试函数进行优化实验分析,结果表明改进的樽海鞘算法相较于基本的樽海鞘算法、其他已经改进的樽海鞘算法在求解精度以及收敛速度上均有明显的提高,且具有更佳的鲁棒性。

Abstract:: In order to improve the shortcomings such as slow convergence speed and poor local optimization ability? of the salp swarm algorithm in solving problems, different optimization strategies are introduced and an improved salp swarm algorithm based on beetle antennae search and adaptive is proposed. The improved salp swarm algorithm introduces the beetle antennae search and Lorentz function instead of the random value in the leader position updating formula of the basic salp algorithm,which improves its local optimi-zation ability. Introducing adaptive inertial weights in the follower’s position update to adjust the influence of its own position and the previous generation’ s position on the follower’s position update, providing a better balance between global search and local search. The improved salp swarm algorithm has been optimized and analyzed by 8 different test functions. It is showed that compared with the basic salp swarm algorithm and other improved salp swarm algorithm, the proposed algorithm has obvious improvement in solving accuracy and convergence speed with better robustness.