双重反馈机制的蚁狮算法.pdfVIP

Computer Engineering and Applications 计算机工程与应用 2017 ，53（12） 31 双重反馈机制的蚁狮算法 吴伟民，张晶晶，林志毅，苏 庆 WU Weimin, ZHANG Jingjing, LIN Zhiyi, SU Qing 广东工业大学 计算机学院，广州 510006 School of Computer, Guangdong University of Technology, Guangzhou 510006, China WU Weimin, ZHANG Jingjing, LIN Zhiyi, et al. Antlion optimization algorithm based on double feedback mecha- nism. Computer Engineering and Applications, 2017, 53 （12）：31-35. Abstract ：The Antlion Optimization （ALO ）algorithm with low convergence precision and easy to fall into the local opti- mizations, the characteristics of the antlions ’ability and the population improvement rate as the double feedback informa- tion are introduced into the ALO algorithm , so the ALO algorithm based on Double Feedback mechanism （DFALO ）is proposed. DFALO algorithm uses dynamic adaptive feedback as adjustment strategy to dynamically adjust the trap size to improve the convergence accuracy. Using spatiotemporal chaos exploration strategy to improve the global search ability, to avoid the algorithm into the local optimal. Using diversity feedback Gaussian mutation strategy to enhance the diversity of the population to avoid the algorithm precocious. Experimental results on eight standard test functions indicate that DFALO has a significant improvement in balance exploration and exploitation, high speed of convergence, strong global search ability and high precision. Key words ：Antlion Optimization algorithm （ALO ）; double feedback; spatiotemporal chaos; Gaussian mutation 摘 要：针对基本蚁狮算法存在的收敛精度低、易陷入局部最优解的缺陷，将蚁狮能力和种群改善率的特征作为双 重反馈信息引入ALO 算法，提出双重反馈机制的蚁狮算法DFALO 。DFALO 算法运用动态自适应反馈调整策略以 动态调整陷阱大小而提高收敛精度；利用时空混沌探索策略提高了全局搜索能力，避免算法陷入局部最优；采用多 样性反馈高斯变异策略增强种群的多样性而避免算法出现早熟。八个标准测试函数仿真测试表明，DFALO 在平衡 全局搜索和局部开发能力上有显著提高，收敛速度快、全局搜索能力强、求解精度高。 关键词：蚁狮算法；双重反馈；时空混沌；高斯变异 文献标志码：A