多策略改进的混合帝国竞争算法

Improved Imperialist Competitive Algorithm based on Quantum Behavior

为了克服帝国竞争算法初始帝国分布不均及易早熟等缺陷,提出一种多策略改进的混合帝国竞争算法。通过拉丁超立方抽样改善由于随机产生的帝国在搜索空间分布不均的状况,以达到扩大算法搜索范围的目的。针对算法后期竞争过程中帝国多样性降低过快而导致易早熟,引入人工蜂群算法中引领蜂与跟随蜂之间的信息反馈机制,形成混合帝国竞争算法。多个测试函数的验证结果表明,改进算法提高了算法寻优精度和全局搜索效率。

In order to overcome the defects of the initial Empire distribution and prematurity in Imperial competition algorithm,a multi strategy improved hybrid Empire competition algorithm is proposed.The purpose of expanding the search scope of the algorithm is to be expanded by the Latin hypercube sampling because the random empires are distributed unevenly in the search space.Aiming at the premature decline of the diversity of the Empire in the later stage of the algorithm,the information feedback mechanism between the bee and the bee was introduced into the artificial bee colony algorithm,and the mixed imperialism competition algorithm was formed.The verification results of multiple test functions show that the improved algorithm improves the precision of optimization and the efficiency of global search.