基于混沌优化技术的并行进化模型及性能分析

PARALLEL EVOLUTIONARY MODEL BASED ON CHAOS OPTIMISATION AND ITS PERFORMANCE ANALYSIS

提出了基于混沌优化技术的并行免疫量子进化算法,算法中将种群分成若干个独立的子群体,称为宇宙。宇宙内采用免疫量子进化算法、灾变算子,宇宙间采用基于学习机制的移民、基于混沌序列的信息交互,因此算法具有更好的种群多样性、更快的收敛速度和全局寻优能力。不仅从理论上分析了算法的性能,而且通过仿真实验验证了该算法的优越性。

A novel parallel immune quantum evolutionary algorithm based on chaotic optimisation technique is proposed. In the algorithm, populations are divided into some independent sub-colonies, and which is named as universes. The immune quantum evolutionary algorithm and catastrophe operator are used inside each universe, the learning mechanism-based emigration and the chaos sequence-based information interaction are adopted among the universes, so that the algorithm has better population diversity, faster convergence speed and global search ability. In this paper the performance of the algorithm is analyzed theoretically, its superiority is also verified by simulation experiments.