论文部分内容阅读
对一个系统进行非线性的判别，进而识别出其非线性的类型，可统称为非线性系统的非参数识别。本文阐述了一些非线性时间序列的非参数识别的方法：对于确定性非线性振动系统，识别方法为极限环法和Ｈｉｌｂｅｒｔ变换；对于随机非线性振动系统，采用概率统计方法．其中有概率密度函数的分布，高阶矩，高阶相关和多维功率谱。计算机仿真实验证明，以上方法是行之有效的。
Non-linear identification of a system, and then identify the type of non-linear, can be collectively referred to as non-linear system of non-parametric identification. This paper describes some non-linear time-series nonparametric identification methods: for deterministic nonlinear vibration system, the identification method is the limit cycle method and the Hilbert transform; for the random non-linear vibration system, using the method of probability and statistics. Among them are the distribution of probability density functions, higher order moments, higher order correlations, and multidimensional power spectra. Computer simulation experiments show that the above method is effective.