基于诺兰模型思想的改进混沌粒子群优化算法及评价

针对粒子群优化算法(PSO) 在处理高维复杂函数时容易陷入局部极值、收敛速度慢的缺陷, 从系统的认知分析过程和角度出发, 提出一种基于诺兰模型(NM) 思想的改进PSO 算法. 该算法在Tent 混沌映射选择的参数的基础上, 结合NM信息融合和协调的思想, 在速度更新过程中增加均衡项, 并设计粒子群的欧氏距离指数以防止早熟, 从而实现对粒子的自动调整、保证多样性和提高算法的全局搜索能力. 最后, 运用典型函数对所提出算法进行测试, 并与最新相关算法进行比较, 结果表明, 所提出算法在全局搜索能力、效率和稳定性方面均具有明显的优势.

Improved chaotic particle swarm optimization algorithm and valuation based on Nolan model thinking

Aiming at the problem that the particle swarm optimization(PSO) algorithm trends to trap in local extreme, and performs high dimensional complex functions inefficiently, from perspectives of the cognitive analysis process of the system, a PSO algorithm based on Nolan model is proposed. The Tent chaotic map is introduced to improve the ergodicity of the algorithm, and the Euclidean distance index is given based on particle average position to automatically adjust particles’ position and ensure diversity to improve the global search capacity of the algorithm. Finally, typical functions are used to test the proposed method. Compared with the current algorithms, it is showed that the proposed method has the advantages in global search, efficiency and stability.