Computer Engineering and Applications ›› 2022, Vol. 58 ›› Issue (19): 174-183.DOI: 10.3778/j.issn.1002-8331.2101-0520

• Pattern Recognition and Artificial Intelligence • Previous Articles     Next Articles

Multi-Strategy Chimp Optimization Algorithm and Its Application of Engineering Problem

HUANG Qian, LIU Sheng, LI Mengmeng, GUO Yuxin   

  1. School of Management, Shanghai University of Engineering Science, Shanghai 201620, China
  • Online:2022-10-01 Published:2022-10-01

多策略黑猩猩优化算法研究及其工程应用

黄倩,刘升,李萌萌,郭雨鑫   

  1. 上海工程技术大学 管理学院,上海 201620

Abstract: Aiming at the problems of chimp optimization algorithm, such as depending on initial population, easily falling into local optimal solution and low convergence accuracy, a multi-strategy chimp optimization algorithm EOSMICOA(chaotic elite opposition-based simple method improved COA) is proposed. In EOSMICOA, Chaotic elite opposition-based learning strategy is used to initialize the individual position of chimpanzee to improve the diversity and quality of the population. Meanwhile, in the process of position updating, simplex method and the individual memory mechanism are used to improve the position of poor individuals, so that can balance the development and exploration ability of the algorithm, and improve the optimization accuracy of the algorithm. In order to verify the optimization ability of the improved algorithm, EOSMICOA is compared with several algorithms on 20 complex functions. The results show that EOSMICOA has obvious advantages in convergence accuracy and optimization speed. Finally, EOSMICOA and several improved algorithms are applied to the optimization of welded beam design. The comparison results show that EOSMICOA can be applied to engineering design optimization problems more effectively.

Key words: chimp optimization algorithm, infinite folding iterative chaotic map, opposition-based learning, simple method, group individual memory mechanism

摘要: 针对基本黑猩猩优化算法存在的依赖初始种群、易陷入局部最优和收敛精度低等问题,提出一种多策略黑猩猩优化算法EOSMICOA(chaotic elite opposition-based simple method improved COA)。在EOSMICOA算法中,利用混沌精英反向学习策略对黑猩猩个体位置进行初始化,提高种群的多样性和质量,同时在位置更新过程中利用单纯形法和群个体记忆机制对较差个体进行改进,进一步提高算法的局部开发能力和勘探能力,以及算法的寻优精度。为验证改进算法的寻优能力,将EOSMICOA算法与多个智能算法对20个复杂函数进行对比实验,结果表明EOSMICOA在收敛精度、寻优速度等方面都有明显优势。最后,将EOSMICOA与当前最新改进算法应用于焊接梁设计中,对比结果表明EOSMICOA可以更有效地应用于工程设计优化问题。

关键词: 黑猩猩优化算法, 无限折叠迭代混沌映射, 精英反向学习, 单纯形法, 群个体记忆机制