[1]毛雪迪,王 冰,夏煌智,等.嵌入 Cat 映射的混合变异探路者算法及其应用[J].计算机技术与发展,2024,34(02):171-179.[doi:10. 3969 / j. issn. 1673-629X. 2024. 02. 025]
　MAO Xue-di,WANG Bing,XIA Huang-zhi,et al.Hybrid Mutation Pathfinder Algorithm Embedded with Cat Mapping and Its Application[J].,2024,34(02):171-179.[doi:10. 3969 / j. issn. 1673-629X. 2024. 02. 025]

点击复制

嵌入 Cat 映射的混合变异探路者算法及其应用()

分享到：

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:

34

期数:

2024年02期

页码:

171-179

栏目:

人工智能

出版日期:

2024-02-10

文章信息/Info

Title:

Hybrid Mutation Pathfinder Algorithm Embedded with Cat Mapping and Its Application

文章编号:

1673-629X(2024)02-0171-09

作者:

毛雪迪; 王 冰; 夏煌智; 张鲁平; 李永超

牡丹江师范学院 数学科学学院,黑龙江 牡丹江 157000

Author(s):

MAO Xue-di; WANG Bing; XIA Huang-zhi; ZHANG Lu-ping; LI Yong-chao

School of Mathematical Sciences,Mudanjiang Normal University,Mudanjiang 157000,China

关键词:

探路者算法; 函数优化问题; Cat 映射; 混合变异; 工程优化问题

Keywords:

pathfinder algorithm; function optimization problem; Cat map; hybrid mutation; engineering optimization problem

分类号:

TP301. 6

DOI:

10. 3969 / j. issn. 1673-629X. 2024. 02. 025

摘要:

针对探路者算法(PFA)求解精度不高、寻优速度较慢与易陷入局部最优等问题,提出一种嵌入 Cat 映射的混合变异探路者算法( CHMPFA) 运用于函数优化问题中。 首先,利用 Cat 混沌映射

的随机性和分散性等特点,再结合反向学习的引导作用,使种群能够覆盖在更为广阔的搜索空间,提高算法的全局搜索能力;其次,在探路者位置更新阶段引入衰减因子平衡算法的全局和局部搜索能力,通过迭代次数的增长逐渐地缩小搜索空间范围,帮助算法快速找到最优解,从而提升算法的搜索效率和收敛速度;最后,利用变异概率随机选择柯西变异或高斯变异对最优个体进行位置扰动,两种变异策略能够帮助个体快速跳出局部最优向其它区域前进。 将 CHMPFA 在 10 个经典基准测试函数和 12 个复杂的 CEC2017 函数集上进行测试,并将其运用于压力容器工程设计问题,与原算法和其它算法的实验结果进行比较,结果表明 CHMPFA 的求解精度、寻优速度与局部最优规避性均明显加强,更低的工程造价成本进一步验证了 CHMPFA 的鲁棒性。

Abstract:

A hybrid mutation pathfinder algorithm embedded with Cat mapping ( CHMPFA) is proposed for the function optimizationproblem in view of the problems of low accuracy of pathfinder algorithm ( PFA) solution,slow speed of finding the best and easy to fallinto local optimum. Firstly, using the characteristics of Cat chaotic mapping such as randomness and dispersion, combined with theguiding effect of opposition-based learning,the population can cover a wider search space and improve the global search capability of thealgorithm. Secondly,the introduction of reduction factors in the pathfinder position update phase balances the global and local search capabilities of the algorithm, gradually narrowing the search space range through the growth of the number of iterations, helping thealgorithm to find the optimal solution quickly,thus enhancing the search speed and convergence of the algorithm. In the end,the optimalindividual is perturbed in position using the mutation probability of randomly selected Cauchy mutation or Gaussian mutation,and the twomutation strategies can help individual quickly jump out of the local optimum to other regions. The CHMPFA is tested on 10 classicalbenchmark test functions and 12 complex CEC2017 function, and applied to pressure vessel engineering design problem. Theexperimental results are compared with those of the original algorithm and other algorithms,and the results show that the CHMPFA is significantly enhanced in terms of solution accuracy, finding speed and local optimum avoidance, and the lower engineering cost furthervalidate the robustness of the CHMPFA.

常用功能