论文部分内容阅读
通过考察现有的多模态优化算法。指出其存在的不足,并根据它们对峰值等高函数搜索效果较好,而对峰值不等高函数效果较差的共同特点,提出评价函数的平衡峰值策略并加以实现.基于免疫系统的抗体进化机制,集成传统的梯度进化思想,设计一种新的多模态免疫算法(MIA).给出算法主要操作算子的具体实现,并分析其运行机理、完全收敛性和计算复杂性.通过仿真实验,验证算法求解多模态问题,特别是求解具有不等高多峰函数的有效性、完全收敛性及快速收敛能力.
By examining the existing multi-modal optimization algorithm. Point out their shortcomings and put forward the balanced peak strategy of evaluation function according to the common features that they search the peak contour function better and the inequality peak inequality effect.From the immune system, Mechanism, integrating the traditional evolutionary gradient idea, a new multi-modal immune algorithm (MIA) is designed.The concrete implementation of the main operators of the algorithm is given, and its operation mechanism, complete convergence and computational complexity are analyzed. Experiments and verification algorithms are used to solve multi-modal problems, especially the validity, complete convergence and fast convergence of unequal multi-modal functions.