多峰函数优化的免疫混沌网络算法

Immune Chaotic Network Algorithm for Multimodal Function Optimization

针对多峰函数优化问题,借鉴混沌遍历特性和免疫网络理论,提出一种免疫混沌网络算法。算法利用混沌运动的自身规律在不同的峰值区域内搜索最佳抗体,增强了算法的局部搜索能力;采用网络抑制策略,保持了种群的多样性;通过网络补充机制自适应地调节抗体群的规模,提高了算法对不同类型多峰函数的适应能力。仿真结果表明该算法能有效地改善种群的多样性,较好地保持全局搜索和局部搜索的动态平衡,具有更强的多峰函数优化能力。

Referred to the ergodicity of chaos and immune network theory,an immune chaotic network algorithm for multimodal function optimization was proposed. The rule of chaotic motion was used to search the best antibodies in different peak regions in order to enhance the capacity of local search. The strategy of network suppression was adopted to maintain the diversity of population. Under the action of network supplement mechanism,the scale of antibody population was adjusted to adapt different types of multimodal function. Simulation results show that the algorithm can not only improve population diversity effectively,but also keep the dynamic balance between global search and local search well. Therefore,it has excellent optimization performance to multimodal function.