论文部分内容阅读
针对混合蛙跳算法在多模态函数寻优中存在的易陷入局部最优、求解精度低、寻优峰值数过少等缺陷,本文提出一种基于圆内衍生变异的免疫双向蛙跳算法.该算法在每次全局循环迭代中,先通过基于双向进化机制的混合蛙跳算法以模因组的形式进行“局部-全局”搜索,再通过双控限幅变异的克隆选择算法对已搜索到的较优解进行局部优化以进一步提高解的精度.最后通过将部分函数旋转以进一步验证算法性能.仿真结果表明,在保证收敛速度的同时,该算法在寻优精度、能搜索到的极值点个数方面均有显著提高。
In order to overcome the shortcoming of hybrid frog leapfrogging algorithm such as local optimal solution, low precision and too few optimal peak values, a new immune two-way frog leapfrog algorithm based on intra-circle derivative mutation is proposed. In each global loop iteration, the algorithm firstly searches for “local-global” by the meso-group by the mixed frog leaping algorithm based on the two-way evolutionary mechanism, and then uses the clonal selection algorithm with double- To optimize the performance of the algorithm.Finally, the partial functions are further rotated to verify the performance of the algorithm.Experimental results show that the proposed algorithm has the advantages of excellent precision, The number of points have significantly increased.