Journal of Frontiers of Computer Science and Technology ›› 2023, Vol. 17 ›› Issue (12): 2896-2912.DOI: 10.3778/j.issn.1673-9418.2210057

• Theory·Algorithm • Previous Articles     Next Articles

Remora Optimization Algorithm Combining Joint Opposite Selection and Host Switching

JIA Heming, WEN Changsheng, WU Di, RAO Honghua, LIU Qingxin, LI Shanglong   

  1. 1. Department of Information Engineering, Sanming University, Sanming, Fujian 365004, China
    2. Department of Education and Music, Sanming University, Sanming, Fujian 365004, China
    3. School of Computer Science and Technology, Hainan University, Haikou 570228, China
  • Online:2023-12-01 Published:2023-12-01

融合联合反向学习与宿主切换机制的䲟鱼优化算法

贾鹤鸣,文昌盛,吴迪,饶洪华,刘庆鑫,力尚龙   

  1. 1. 三明学院 信息工程学院,福建 三明 365004
    2. 三明学院 教育与音乐学院,福建 三明 365004
    3. 海南大学 计算机科学与技术学院,海口 570228

Abstract: The remora optimization algorithm (ROA) is a meta heuristic optimization algorithm proposed in 2021. It simulates the behavior of parasitic attachment to the host, empirical attack and host foraging in the ocean. The structure of ROA is simple and easy to implement, but the overall situation is slightly insufficient, which easily leads to ROA’s slow convergence speed and even difficult convergence in the later period. To solve the above problems, host switching mechanism is added in the exploration phase, and new host beluga is introduced to improve the exploration ability of original ROA. At the same time, through adding joint opposite selection strategy, the ability of the algorithm to jump out of the local optimum is enhanced, and the comprehensive optimization performance of the algorithm is further improved. Through the above improvements, an improved remora optim-ization algorithm (IROA) is proposed, which integrates the joint opposite selection and host switching mechanism. In order to verify the performance and improvement advantages of IROA, IROA is compared with the original ROA, six typical original algorithms and four improved algorithms on ROA. Experimental results of CEC2020 standard test function show that IROA has stronger optimization ability and higher convergence accuracy. Finally, the advantages and engineering applicability of the improved algorithm are further verified by solving the car crashworthiness design problem.

Key words: remora optimization algorithm, meta heuristic optimization algorithm, joint opposite selection, host switching mechanism, beluga whale optimization, benchmark function test, engineering problem solving

摘要: 䲟鱼优化算法(ROA)是2021年提出的元启发式优化算法,其模拟了海洋中?鱼寄生依附宿主、经验攻击和宿主觅食的行为。ROA的结构简单且易于实现,但全局性稍显不足,易导致算法收敛速度慢甚至后期难以收敛的现象。针对上述问题,在探索阶段加入宿主切换机制,引入新宿主白鲸,提高原算法的探索能力;同时加入联合反向学习策略,增强了算法跳出局部最优的能力,进一步提高了算法的综合优化性能。通过以上改进,提出了一种融合联合反向学习与宿主切换机制的?鱼优化算法(IROA)。为了验证IROA的性能与改进优势,将IROA与原始ROA、6种典型的原始算法以及4种关于ROA的改进算法进行对比。通过CEC2020标准测试函数的实验结果表明,IROA具有更强的寻优能力和更高的收敛精度;最后针对汽车防撞性设计问题的求解,进一步验证了IROA的优势和工程适用性。

关键词: ?鱼优化算法, 元启发式优化算法, 联合反向学习, 宿主切换机制, 白鲸优化算法, 基准函数测试, 工程问题求解