Ouyang Chengtian, Liao Chang, Zhu Donglin, Zheng Yangyang, Zhou Changjun, Li Taiyong
School of Information Engineering, Jiangxi University of Science and Technology, Ganzhou, 341000, China.
School of Computer Science and Technology, Zhejiang Normal University, Jinhua, 321004, China.
Sci Rep. 2024 Mar 28;14(1):7445. doi: 10.1038/s41598-024-58029-3.
The original Harris hawks optimization (HHO) algorithm has the problems of unstable optimization effect and easy to fall into stagnation. However, most of the improved HHO algorithms can not effectively improve the ability of the algorithm to jump out of the local optimum. In this regard, an integrated improved HHO (IIHHO) algorithm is proposed. Firstly, the linear transformation escape energy used by the original HHO algorithm is relatively simple and lacks the escape law of the prey in the actual nature. Therefore, intermittent energy regulator is introduced to adjust the energy of Harris hawks, which is conducive to improving the local search ability of the algorithm while restoring the prey's rest mechanism; Secondly, to adjust the uncertainty of random vector, a more regular vector change mechanism is used instead, and the attenuation vector is obtained by modifying the composite function. Finally, the search scope of Levy flight is further clarified, which is conducive to the algorithm jumping out of the local optimum. Finally, in order to modify the calculation limitations caused by the fixed step size, Cardano formula function is introduced to adjust the step size setting and improve the accuracy of the algorithm. First, the performance of IIHHO algorithm is analyzed on the Computational Experimental Competition 2013 (CEC 2013) function test set and compared with seven improved evolutionary algorithms, and the convergence value of the iterative curve obtained is better than most of the improved algorithms, verifying the effectiveness of the proposed IIHHO algorithm. Second, the IIHHO is compared with another three state of the art (SOTA) algorithms with the Computational Experimental Competition 2022 (CEC 2022) function test set, the experiments show that the proposed IIHHO algorithm still has a strong ability to search for the optimal value. Third, IIHHO algorithm is applied in two different engineering experiments. The calculation results of minimum cost prove that IIHHO algorithm has certain advantages in dealing with the problem of search space. All these demonstrate that the proposed IIHHO is promising for numeric optimization and engineering applications.
原始的哈里斯鹰优化（HHO）算法存在优化效果不稳定且易陷入停滞的问题。然而，大多数改进的HHO算法无法有效提高算法跳出局部最优的能力。对此，提出了一种集成改进的HHO（IIHHO）算法。首先，原始HHO算法使用的线性变换逃逸能量相对简单，缺乏实际自然界中猎物的逃逸规律。因此，引入间歇能量调节器来调整哈里斯鹰的能量，这有利于在恢复猎物休息机制的同时提高算法的局部搜索能力；其次，为调整随机向量的不确定性，改用更规则的向量变化机制，并通过修改复合函数得到衰减向量。最后，进一步明确了莱维飞行的搜索范围，有利于算法跳出局部最优。最后，为修正固定步长带来的计算局限性，引入卡尔达诺公式函数来调整步长设置，提高算法的精度。首先，在2013年计算实验竞赛（CEC 2013）函数测试集上分析IIHHO算法的性能，并与七种改进的进化算法进行比较，得到的迭代曲线收敛值优于大多数改进算法，验证了所提出的IIHHO算法的有效性。其次，使用2022年计算实验竞赛（CEC 2022）函数测试集将IIHHO与另外三种先进（SOTA）算法进行比较，实验表明所提出的IIHHO算法仍具有很强的搜索最优值的能力。第三，将IIHHO算法应用于两个不同的工程实验。最小成本的计算结果证明IIHHO算法在处理搜索空间问题方面具有一定优势。所有这些都表明所提出的IIHHO在数值优化和工程应用方面具有广阔前景。
