Wang Zhengzong, Ye Xiantao, Jiang Guolin, Yi Yiru
School of Intelligent Manufacturing and Electronic Engineering, Wenzhou University of Technology, Wenzhou 325035, China.
Zhejiang Zhengli Enterprise Management Co., Ltd., Wenzhou 325035, China.
Biomimetics (Basel). 2025 Jun 1;10(6):354. doi: 10.3390/biomimetics10060354.
In order to overcome the inherent drawbacks of the baseline Zebra Optimization Algorithm (ZOA) approach, such as its propensity for premature convergence and local optima trapping, this work creates a Multi-Strategy Enhanced Zebra Optimization Algorithm (MZOA). Three strategic changes are incorporated into the improved framework: triangular walk operators to balance localized exploitation and global exploration across optimization phases; Levy flight mechanisms to strengthen solution space traversal capabilities; and lens imaging inversion learning to improve population diversity and avoid local convergence stagnation. The enhanced solution accuracy of the MZOA over modern metaheuristics is empirically validated using the CEC2005 and CEC2017 benchmark suites. The proposed MZOA's performance improved by 15.8% compared to the basic ZOA The algorithm's practical effectiveness across a range of environmental difficulties is confirmed by extensive assessment in engineering optimization and robotic route planning scenarios. It routinely achieves optimal solutions in both simple and complicated setups. In robot path planning, the proposed MZOA reduces the movement path by 8.7% compared to the basic ZOA. These comprehensive evaluations establish the MZOA as a robust computational algorithm for complex optimization challenges, demonstrating enhanced convergence characteristics and operational reliability in synthetic and real-world applications.
为了克服基线斑马优化算法（ZOA）方法的固有缺点，例如其过早收敛和陷入局部最优的倾向，本文提出了一种多策略增强斑马优化算法（MZOA）。改进后的框架纳入了三项策略性改变：三角游走算子，用于在优化阶段平衡局部开发和全局探索；莱维飞行机制，以增强解空间遍历能力；以及透镜成像反演学习，以提高种群多样性并避免局部收敛停滞。使用CEC2005和CEC2017基准测试套件，通过实验验证了MZOA相对于现代元启发式算法的增强求解精度。与基本ZOA相比，所提出的MZOA的性能提高了15.8%。通过在工程优化和机器人路径规划场景中的广泛评估，证实了该算法在一系列环境困难下的实际有效性。它在简单和复杂设置中都能常规地获得最优解。在机器人路径规划中，与基本ZOA相比,所提出的MZOA将移动路径减少了8.7%。这些综合评估将MZOA确立为一种用于复杂优化挑战的强大计算算法，在合成和实际应用中展示了增强的收敛特性和操作可靠性。
