摘要
为了对分数阶超混沌系统中的未知参数进行准确估计,提出一种量子混沌粒子群优化算法(Quantum chaos particle swarm optimization,QCPSO).该算法通过对量子粒子群优化算法(Quantum behaved particle swarm optimization,QPSO)的实现机理进行分析,并结合量子纠缠与混沌系统之间的相关性而实现.首先,将量子势阱中心视为混沌吸引子围绕的不动点,处于吸引子外部的粒子会逐渐聚集于吸引子之内,而处于吸引子内部的粒子会出现快速分离扩散的现象;然后,采用基于随机映射的粒子更新机制,充分保证混沌粒子的初值多样性;最后,提出了基于不动点中心的尺度自适应策略,解决了算法后期的搜索停滞问题.运用QCPSO算法对典型分数阶超混沌系统参数进行估计,结果表明,该算法在收敛速度与精度上优于改进的差分进化算法、自适应人工蜂群算法以及改进的量子粒子群优化算法.
A new quantum chaos particle swarm optimization （QCPSO） was proposed to accurately estimate the un- certain parameters of the fractional order hyper chaotic system. The QCPSO algorithm was realized by analyzing the mecha- nism of quantum behaved particle swarm optimization （ QPSO ） and combining the correlation between quantum entangle- ment and chaotic system. Firstly, the center of potential well was replaced by a fixed point of chaotic attractor. The particles which outside the attractor were gradually converged to the attractor, and the particles which inside the attractor were quickly diffused. Secondly, in order to guarantee the diversity of the initial value of the chaotic particles, the particle update mechanism based on random mapping was proposed. Finally, a scale adaptive strategy was proposed to solve the problem of search stagnation of the algorithm. The parameters of fractional order hyper chaotic system were estimated by the QCPSO algo- rithm, and the results showed that the QCPSO algorithm has faster convergence speed and higher accuracy than improved differential evolution algorithm, adaptive artificial bee colony algorithm and improved QPSO algorithm.
出处
《电子学报》
EI
CAS
CSCD
北大核心
2018年第2期333-340,共8页
Acta Electronica Sinica
基金
中国科学院西部之光基金(No.2011180)
关键词
量子粒子群优化算法
混沌映射
混沌吸引子
分数阶超混沌系统
quantum behaved particle swarm optimization
chaotic maps
strange attractor
fractional order hyper chaotic system
