【摘要】 针对已有协同进化算法种群分割方法的不足,借鉴连接识别思想,提出用分段间隔相关迭代连接识别进行种群分割的算法,并给出算法的思想、步骤及具体实施方案.将其应用于典型的复杂函数优化问题,简要分析算法在搜索效率、计算复杂性等方面的性能以及相关参数对算法性能的影响,说明算法的可行性.更多还原

【Abstract】 <Abstrcat> Based on the idea of linkage identification, piecewise interval correlation by iteration (PICI) linkage identification based population decomposition algorithm is presented in allusion to the deficiency of existing population decomposition methods in coevolutionary algorithm. The methodology, steps and strategies of population decomposition are given in detail. The performance in search efficiency, computation complexity of the algorithm is analyzed by optimizing benchmark complicated functions and the effect of pertinent parameters on its performance is also analyzed.更多还原