混沌序列构造方法研究  

A Producing Method of Chaos

作　　者：肖坤[1] 宋恒[1] 王振家[1] 

机构地区：[1]空军工程大学,陕西西安710038

出　　处：《计算机仿真》2005年第1期187-190,共4页Computer Simulation

摘　　要：针对扩频通信中运用神经网络方法产生的混沌序列性能难以调整的问题 ,提出并设计由遗传算法搜索调整系统适配参数的神经模糊推理系统 (ANFIS)生成混沌序列。在该系统训练的过程中 ,将其产生的序列的Lyapunov指数与平衡性、自(互 )相关性、数量的加权平均函数作为适应度函数 ,通过遗传算法调整系统参数 ,产生出符合扩频通信应用要求的混沌序列。通过计算机仿真试验 ,并与Logistic混沌映射 (教师值 )相比较 ,结果表明该方法能产生出随机性能良好的混沌序列 。According to the problem that the chaotic sequences generated by neural network are difficult to adjust, this paper presents a Chaos Generation ANFIS model whose basic system parameters are adjusted by genetic algorithm. In the course of its training, which forms its adaption function with the Lyapunov index,balance, self-relativity, mutual-relativity, quantity of chaos sequences generated by itself, adjusted by genetic algorithm, it generates proper chaos sequences. Simulation has shown this model can generate a chaotic sequence whose randomness is excellent. Changing the output sequence is intelligent by adjusting genetic algorithm. And the model's sensitivity to initial values is very high, which is suitable for multi-user's mobile communication in CDMA system.

关 键 词：混沌 相空间重构 神经模糊推理系统 遗传算法 扩频通信 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]