基于遗传扰动机制的改进蝙蝠优化算法

An improved bat optimization algorithm based on genetic disturbance mechanism

针对蝙蝠算法现存的缺点,如收敛速度慢、优化精确度低、早熟,提出一种基于遗传扰动机制的改进蝙蝠算法(GDBA).该优化算法引入了遗传竞争机制,通过比较与全局最优解的差异,随时调整遗传算法的交叉率和变异率,使得种群具有遗传性和多样性,解决了蝙蝠算法早熟的问题,同时加快了收敛速度,提高了优化精度.采用基准测试函数进行仿真验证,实验结果表明:与蝙蝠算法(BA)和基于速度权重扰动机制的改进蝙蝠算法(WDBA)相比,该算法(GDBA)具有更好的收敛速度和搜索精度,加强了寻找全局最优解的能力.

Aimed at the defect existing in bat algorithm such as slow convergence,low optimization accuracy,and premature convergence,an improved bat algorithm (GDBA) is proposed based on genetic disturbance mechanism.A genetic competitive mechanism is introduced into this optimization algorithm,its crossing and mutation probabilities are adjusted at any time by means of comparing the difference between its solution and the global optimal solution,making the population to possess heritability and diversity,resolving the prematurity of the bat algorithm,and at the same time accelerating the convergence speed and improving the optimizing precision.The standard test functions are used to conduct simulative verification and its test result shows that compared with the bat algorithm and an improved bat optimization algorithm based on velocity-weighted disturbance mechanism,the GDBA will have faster convergence speed and searching accuracy and it will enhance the ability to search the global optimum solution.