论文部分内容阅读
借鉴蚁群分工的原理,模仿蚂蚁影响其他蚂蚁的刺激行为,提出了一种基于蚁群分工的自适应差分进化算法(DE-Dol,Division of Labor in Differential Evolution)。将个体选择了成功或者失败的策略分别看作是此策略对这个个体的正、负刺激。在该算法中,对于在当前种群中的每个目标向量,根据它对前几代中产生的改善的解决方案的刺激,从策略候选池中选择一个实验向量。这样,3个控制参数(缩放因子F,交叉概率CR和种群数量NP)和变异策略都是根据与策略相关的正、负刺激逐渐自适应的。随后,将DE-Dol算法通过25组CEC2005标准测试函数进行测试。实验结果验证了该方法的有效性和实用性。
By using the principle of ant colony division and imitating the ants’ influence on the stimuli of other ants, a DE-Dol, Division of Labor in Differential Evolution (DE-DOD) algorithm is proposed. The individual’s strategy of choosing success or failure, respectively, as the positive and negative stimulus to this individual. In this algorithm, for each target vector in the current population, one experiment vector is selected from the pool of strategy candidates based on its stimulation of the improved solution generated in previous generations. In this way, the three control parameters (scaling factor F, crossover probability CR and population number NP) and mutation strategy are gradually adaptive based on positive and negative stimuli related to the strategy. Subsequently, the DE-Dol algorithm was tested with 25 sets of CEC2005 standard test functions. The experimental results verify the effectiveness and practicability of this method.