由于人工蜂群（artificial bee colony ，ABC）算法存在收敛速度慢、易陷入局部最优的缺点，采用设置自适应缩放因子和基于适应度排序的选择方式代替传统的轮盘赌模型，提出了一种改进的快速人工蜂群算法（fast artifi-cial bee colony ，FABC）。基于这种FABC算法对4个离散变量的几何优化模型进行了优化，并与遗传算法（GA）、蚁群算法（ACA）、启发式粒子群优化算法（HPSO）和群搜索算法（GSO）作了比较。结果表明，这种改进的人工蜂群算法具有较好的收敛精度。另外，ABC算法以及FABC算法结构简单，可应用在其他优化问题上。%Since the artificial bee colony (ABC) algorithm had drawbacks of slow convergence and easy to fall into local optimum ,a modified optimization algorithm called fast artificial bee colony (FABC) was introduced by setting the self-adaptive scaling factor and replacing the traditional roulette wheel selection model with the fitness sorting method .The suitability of FABC for ge-ometry optimization of structures was tested on four truss examples with discrete variables ,com-pared with other optimal algorithms of GA ,ACA , HPSO and GSO . The results show that FABC had advantages over other algorithms in convergence accuracy .Furthermore ,due to the sim-ple algorithm structure ,both algorithms of ABC and FABC can apply to solving other optimal issues .
展开▼
