中文摘要：

针对混沌系统未知参数的辨识问题,结合人工蜂群搜索算子和混沌优化策略,提出一种自适应混合引力搜索算法,并应用于混沌系统未知参数的优化辨识.利用混沌序列初始化种群以增强搜索初期的遍历性,基于人工蜂群搜索算子进行变异操作以提高算法的局部寻优能力,依据粒子的性能对进化过程中的万有引力系数进行自适应调整,有效避免了早熟收敛,提高了算法的整体寻优性能.以测试函数和典型混沌系统为例进行仿真实验,结果证明该算法具有良好的全局探测和局部开发能力,与遗传算法、粒子群算法、量子粒子群算法和引力搜索算法比较,其对混沌系统参数的估计具有相对较高的辨识精度和收敛速度,算法的有效性得到了验证.

英文摘要：

In order to solve the identification of unknown parameters of a chaotic system, an adaptive hybrid gravitational search algorithm was proposed and applied to estimate the unknown parameters of chaotic systems by reference to the artificial bee colony search operator and chaos optimization strategy. The chaotic sequence was used to build an initial population for enhancing the ergodicity of early search, a mutation operation based on the artificial bee colony search operator was used for improving the ability of local optimization algorithms, and the gravitational coefficients was adjusted adaptively according to the performance of particles; therefore the prema~tre convergence was avoided effectively and the overall performance of the algorithm was improved. Numerical simulations about the test functions and typical chaotic system were conducted, with the results showing that the proposed algorithm has good global detection and local development capabilities and, compared with genetic algorithms, particle swarm optimization, quantum particle swarm optimization and gravitational search algorithm, the proposed algorithm also has a relatively higher recognition accuracy and convergence rate dttring the parameter identification of the chaotic system, and the effectiveness of the proposed algorithm has been verified.