Yang Xiaorui, Zhang Yumei, Lv Xiaojiao, Yang Honghong, Sun Zengguo, Wu Xiaojun
School of Computer Science, Shaanxi Normal University, Xi'an, China.
Key Laboratory of Intelligent Computing and Service Technology for Folk Song, Xi'an, China.
Math Biosci Eng. 2023 May 19;20(7):12263-12297. doi: 10.3934/mbe.2023546.
To address the problems of slow convergence speed and low accuracy of the chimp optimization algorithm (ChOA), and to prevent falling into the local optimum, a chaos somersault foraging ChOA (CSFChOA) is proposed. First, the cat chaotic sequence is introduced to generate the initial solutions, and then opposition-based learning is used to select better solutions to form the initial population, which can ensure the diversity of the algorithm at the beginning and improve the convergence speed and optimum searching accuracy. Considering that the algorithm is likely to fall into local optimum in the final stage, by taking the optimal solution as the pivot, chimps with better adaptation at the mirror image position replace chimps from the original population using the somersault foraging strategy, which can increase the population diversity and expand the search scope. The optimization search tests were performed on 23 standard test functions and CEC2019 test functions, and the Wilcoxon rank sum test was used for statistical analysis. The CSFChOA was compared with the ChOA and other improved intelligent optimization algorithms. The experimental results show that the CSFChOA outperforms most of the other algorithms in terms of mean and standard deviation, which indicates that the CSFChOA performs well in terms of the convergence accuracy, convergence speed and robustness of global optimization in both low-dimensional and high-dimensional experiments. Finally, through the test and analysis comparison of two complex engineering design problems, the CSFChOA was shown to outperform other algorithms in terms of optimal cost. For the design of the speed reducer, the performance of the CSFChOA is 100% better than other algorithms in terms of optimal cost; and, for the design of a three-bar truss, the performance of the CSFChOA is 6.77% better than other algorithms in terms of optimal cost, which verifies the feasibility, applicability and superiority of the CSFChOA in practical engineering problems.
为了解决黑猩猩优化算法（ChOA）收敛速度慢和精度低的问题，并防止陷入局部最优，提出了一种混沌翻腾觅食黑猩猩优化算法（CSFChOA）。首先，引入猫混沌序列生成初始解，然后采用基于对立学习的方法选择更好的解来形成初始种群，这可以在算法开始时确保多样性，提高收敛速度和最优搜索精度。考虑到算法在最后阶段可能会陷入局部最优，以最优解为支点，采用翻腾觅食策略，用镜像位置适应度更好的黑猩猩替换原始种群中的黑猩猩，这可以增加种群多样性，扩大搜索范围。对23个标准测试函数和CEC2019测试函数进行了优化搜索测试，并使用Wilcoxon秩和检验进行统计分析。将CSFChOA与ChOA和其他改进的智能优化算法进行了比较。实验结果表明，CSFChOA在均值和标准差方面优于大多数其他算法，这表明CSFChOA在低维和高维实验中的收敛精度、收敛速度和全局优化鲁棒性方面表现良好。最后，通过对两个复杂工程设计问题的测试和分析比较，表明CSFChOA在最优成本方面优于其他算法。对于减速器设计，CSFChOA在最优成本方面的性能比其他算法好100%；对于三杆桁架设计，CSFChOA在最优成本方面的性能比其他算法好6.77%，这验证了CSFChOA在实际工程问题中的可行性、适用性和优越性。
