融合蜂群行为的量子进化算法

Bee-Behaved Colony Quantum-Inspired Evolutionary Algorithm

为提高量子进化算法的收敛精度和收敛速度,以人工蜂群算法为基本进化框架,提出一种融合蜂群行为的量子进化算法.将采用相位编码的量子进化种群划分为量子开采种群、量子跟随种群以及量子侦察种群,在每个种群内模拟蜜蜂觅食行为寻优,其中量子开采种群采用混沌扰动搜索,量子跟随种群采用柯西变异操作进化.同时对所有种群个体采用量子染色体的两步旋转更新方法,并进行自适应的动态变异操作.利用基准测试函数进行仿真,与相关方法对比分析可知,所提出的算法在大部分的函数上都表现出较好的性能,能有效提高全局收敛性能.

In order to promote the convergence precision and speed of quantmu-inspired evolutionary algorithm, a new bee-behaved quantum-inspired evolutionary algorithm is proposed based on the framework of ABC algorithm The whole population can be encoded with phase and can be divided into three populations ,which are named as quantum employed population, quantum onlooker population and quantum scout population. Every sub-population can work in term of bee behaviors, quantum employed population perform the chaos search and the quantum onlooker population can perform the Cauchy mutation. Every individual in the population can be rotated in two steps, and dynamic mutation operation can also act on every individual. Simulation results of benchmark functions show that the proposed algorithm performs well on most of functions and can get better convergence results.