具有协同寻优的蝙蝠萤火虫混合优化算法

The Hybrid Bat and Firefly Algorithm with Collaborative Optimization

关键词:  函数优化  蝙蝠算法  萤火虫算法  协同寻优

Key words:  function optimization  bat algorithm  firefly algorithm  collaborative optimization

摘要:

萤火虫算法存在着对于初始解分布的依赖性、后期收敛速度慢、易于停滞、早熟和求解精度低等缺陷。本研究在萤火虫算法引入蝙蝠种群在全局最优附近进行更加详细的局部搜索，以协助萤火虫种群进行寻优；并在寻优过程中加强蝙蝠种群与萤火虫种群的信息交互，协调寻优；最后对全局最优个体进行高斯扰动以增加种群的多样性，从而避免种群陷入局部最优解。通过使用6个常见的基准测试函数对该算法进行测试，并与其他3种算法（标准粒子群算法、蝙蝠算法、萤火虫算法）进行对比实验，结果表明该混合算法的总体性能优于其他3种算法。引入蝙蝠种群对萤火虫性能有较大提升，改善切实有效。

Abstract:

The improved firefly algorithm has defects such as dependence on the initial solution distribution, slow convergence in the later stage, easy stagnation, early maturity and low accuracy. This algorithm introduces the bat population to do a more detailed local search near the global optimum to help the firefly population to optimize. In the process of optimization, the information interaction between the bat population and the firefly population is enhanced, and the optimization is coordinated. Finally, Gaussian perturbation is applied to the global optimum individuals to increase the diversity of the population and avoid the population falling into the local optimum solution. The algorithm was tested by using six benchmark functions and compared with the other three algorithms (standard particle swarm algorithm, bat algorithm, firefly algorithm).The results show the overall performance of this hybrid algorithm is better than the other three algorithms. The introduction of the bat population has greatly improved the performance of firefly, and the improvement is effective.