改进基于记忆的人工蜂群算法
发布时间：2019-07-19 08:04
【摘要】：基于记忆的人工蜂群算法(ABCM)通过记住成功使用的邻居和系数指导人工蜂群下一步的搜索,需消耗多次函数评价收敛到吸引子,且始终使用与上次相同的排斥系数,造成收敛速度不快、多样性不足,易陷入局部最优解.提出一种改进ABCM(IABCM),当使用吸引系数时,候选解只消耗一次函数评价收敛到吸引子,如果候选解好于当前解,则替换当前解,否则直接删除该记忆,这样可以利用尽量小的代价得到尽量大的收益.当使用排斥系数时,该系数的数值部分重新随机生成,以增加多样性和随机性,有利于算法跳出局部最优解.在22个不同类型函数上的实验表明,IABCM在收敛速度和精度方面明显优于ABCM.
[Abstract]:The memory-based artificial beehive algorithm (ABCM) guides the next search of the artificial bee colony by remembering the neighbors and coefficients successfully used. It needs to consume many times of function evaluation to converge to the attractor, and always use the same rejection coefficient as the last time, resulting in the convergence speed is not fast, the diversity is insufficient, and it is easy to fall into the local optimal solution. In this paper, an improved ABCM (IABCM), is proposed, when the attraction coefficient is used, the candidate solution only consumes one time to converge to the attractor. If the candidate solution is better than the current solution, the current solution is replaced, otherwise the memory is deleted directly, so that the maximum benefit can be obtained by using the smallest cost. When the rejection coefficient is used, the numerical part of the coefficient is regenerated randomly to increase the diversity and randomness, which is helpful for the algorithm to jump out of the local optimal solution. The experiments on 22 different types of functions show that IABCM is obviously superior to ABCM. in convergence speed and accuracy.
【作者单位】： 韩山师范学院计算机与信息工程学院;上海海事大学信息工程学院;
【基金】：国家自然科学基金项目(61672338,61373028)
【分类号】：TP18
，
本文编号：2516183
[Abstract]:The memory-based artificial beehive algorithm (ABCM) guides the next search of the artificial bee colony by remembering the neighbors and coefficients successfully used. It needs to consume many times of function evaluation to converge to the attractor, and always use the same rejection coefficient as the last time, resulting in the convergence speed is not fast, the diversity is insufficient, and it is easy to fall into the local optimal solution. In this paper, an improved ABCM (IABCM), is proposed, when the attraction coefficient is used, the candidate solution only consumes one time to converge to the attractor. If the candidate solution is better than the current solution, the current solution is replaced, otherwise the memory is deleted directly, so that the maximum benefit can be obtained by using the smallest cost. When the rejection coefficient is used, the numerical part of the coefficient is regenerated randomly to increase the diversity and randomness, which is helpful for the algorithm to jump out of the local optimal solution. The experiments on 22 different types of functions show that IABCM is obviously superior to ABCM. in convergence speed and accuracy.
【作者单位】： 韩山师范学院计算机与信息工程学院;上海海事大学信息工程学院;
【基金】：国家自然科学基金项目(61672338,61373028)
【分类号】：TP18
，
本文编号：2516183
本文链接：https://www.wllwen.com/kejilunwen/zidonghuakongzhilunwen/2516183.html
