从实测时间序列中构造混沌系统的模型是非线性时间序列分析中的一个重要议题,本文利用遗传编程方法(GP),在尽可能大的函数空间内搜索合适的模型结构,并引入粒子群算法(PSO)在动态模型结构情况下进行非线性参数估计,取得了较好效果.此外,演化建模的实现结合了非线性时间序列分析(NTSA)的结果,以NTSA的结果指导演化建模参数的选取并作为模型优劣的评判标准,改进了经典GP算法对混沌系统建模的应用效果.

This paper proposes Genetic Programming Modeling (GPM) algorithm on chaotic time series.GP is used here to search for appropriate model structures in function space,and Particle Swarm Optimization (PSO) algorithm is introduced for Nonlinear Parameter Estimation (NPE) of dynamic model structures.In addition,GPM integrates the results from Nonlinear Time Series Analysis (NTSA) to adjust the parameters and as the criterion of founded models.The simulation shows the effectiveness of such improvements on modeling chaotic time series.