人类行为和t-分布变异策略改进的鸽群优化算法
Improved Pigeon-inspired Optimization Algorithm Based on Human Behavior and t-distributed Mutation Strategy
倪云峰 1梁佳豪 1王静 1郭苹1
作者信息
- 1. 西安科技大学通信与信息工程学院 西安 710600
- 折叠
摘要
针对鸽群优化算法(PIO)全局搜索能力弱、容易陷入局部最优解的缺点,提出一种基于人类行为和t-分布变异策略改进的鸽群优化算法,采用人类行为策略对鸽群优化算法中的地图和指南针算子进行改进,实现种群分布的多样性,采用t-分布变异策略对鸽群优化算法中的地标算子进行改进,提升算法的探索和开发能力,综合两种改进方法提升了算法跳出局部最优解的能力和全局搜索精度.在11个测试函数上与经典的PIO、其他五种算法以及其他学者改进的算法进行对比,实验结果表明,综合两种算法改进的鸽群优化算法具有更优的收敛精度和更快的收敛速度.
Abstract
To address the shortcomings of the pigeon-inspired optimization algorithm(PIO)with weak global search ability and falling into local optimal solutions easily,this paper proposes an improved pigeon-inspired optimization algorithm,which com-bines human behavior strategy and t-distribution mutation strategy.The map compass operator is improved by human behavior strate-gy to realize the diversity of population distribution,and the landmark operator is improved by t-distribution mutation strategy to im-prove the exploration and development capability of the algorithm.Combining the two strategies,the improved pigeon-inspired opti-mization algorithm has stronger ability to jump out of local extremum and higher global searcher accuracy.Comparing with the pi-geon-inspired optimization algorithm(PIO),other five algorithms and other scholars improved PIO on 11 test functions,the experi-mental results show that the improved pigeon-inspired optimization algorithm with two improved strategies has better convergence precision and faster convergence speed.
关键词
改进鸽群搜索算子/人类学习行为/t-分布/全局搜索/局部寻优Key words
improved pigeon swarm search operator/human behavior/t-distribution/global searching/local searching引用本文复制引用
出版年
2024
国家科技期刊平台
NSTL
万方数据