中文摘要：

群搜索优化算法（GroupSearchOptimizer，GSO）具有广泛的生物学背景，特别是引入动物的视觉搜索机制，并且同一些已有的群智能算法相比较，在高维多峰问题上有更好的效果。但算法在个体觅食策略的选择上以及整个动物群体间信息共享的网络拓扑结构来看，存在错过最优值和信息交流模式过于简单的缺陷。受NW模型的启发，同时采用动态采样的方式提出了交互变邻域微分进化群搜索优化算法（Interactive Dynamic Neighborhood Differential EvolutionaryGSO，IDGSO），并采用均匀设计和线性回归方法对参数进行选择，4个标准测试函数表明了IDGSO的有效性。

英文摘要：

Group Search Optimizer （GSO） has the advantage of the design from a biological view, while animal scanning mechanisms are employed metaphorically to design optimum searching strategies for solving continuous optimization problems. Compared with some existing group intelligence algorithms, it has a better effect in the high dimensional problems. But from the individual foraging strategies it choose and the entire animal groups information sharing network topology, global optimal possibly missed and information exchange model is simple. Inspiration from the Newman and Watts model, Interactive Dynamic Neighborhood GSO （ IDGSO ） is proposed based on dynamic sampling. Adopting uniform designand the linear regression method on the parameter selection, 4 benchmark functions demonstrate the effectiveness of the algorithm.