一种改进的克隆选择算法在多峰值函数优化中的应用

为了解决deCastro2002年提出的CLONALG算法在多峰值函数优化时多峰搜索能力弱、训练时间长的问题，本文提出了一种改进的克隆选择算法。该算法运用新的克隆选择操作、克隆变异操作和最佳抗体停止进化操作，并且引入了抗体抑制操作，不仅可以动态调整种群大小，具有较强的全局和局部搜索能力，而且搜索时间较短。与Castro的克隆选
择算法相比，本文算法在较短的时间内可以搜索到全局最优解和更多的局部最优解。

An Improved Clone Selection Algorithm for Multi-Model Function Optimization

To overcome the drawbacks of the CLONALG algorithm for multi-model function optimization by de Castro in 2002,an improved clone selection algorithm is proposed in this paper.The algorithm uses new operations of clone selection,clone mutation and the best antibody evolution stop,and adopts the operation of antibody suppression.The algorithm can not only dynamically adjust the population size,but feature strong abilities of local and global search,and the search time is relatively short.Compared with the de Castro clone selection algorithm,the algorithm presented in this paper may search and get the global optimum solution and more local optimum solutions in a shorter time.