基于混沌策略状态转移算法的混沌系统参数辨识*

针对混沌系统的参数辨识是一个多维参数的优化问题,提出了基于混沌策略状态转移算法的混沌系统参数辨识方法。该方法是在初始化时以混沌序列初始化种群,在搜索过程中引入混沌变异机制,利用遍历性对状态进行变异操作,避免了过早收敛,提高了全局搜索能力。利用该算法辨识Lorenz混沌系统参数,并与基本状态转移算法和粒子群算法进行比较。仿真结果表明,在有无噪声干扰的情况下,该算法比粒子群算法和基本状态转移算法具有更好的辨识精度且比粒子群算法具有更好的收敛速度,证明了该算法的有效性和抗干扰性,对混沌理论的发展有重要的意义。

Parameter identification in chaotic systems based on state transition algorithm combined with the chaotic strategy

This paper proposed a chaotic system identification method based on state transition algorithm with the combination of the chaotic strategy (CSSTA) for multidimensional parameter identification of chaotic system. The chaotic strategy initialized the population with chaotic sequences during initialization, and used chaos mutation mechanism in the search process. The ergodicity of the chaotic mutation operator was used to finish the mutation operation of the selected individual. These methods avoided the premature convergence, and improved the global convergence ability of the algorithm. CSSTA was used on identification of Lorenz chaotic system, and compared with particle swarm optimization (PSO) and state transition algorithm (STA). Simulation experiments showed that the proposed method have better identification precision than the particle swarm optimization and state transition algorithm and better convergence than PSO and STA under no noise interference or noise environment. And where there is noise case, state transition algorithm with the combination of chaotic strategy not only has high identification accuracy and convergence, and has a good anti-jamming, which has important significance to the development of chaos theory.