论文部分内容阅读
在介绍了一类输出只与其输入的平方有关的偶型H模型之后,分析了现有相关算法无法估计此类模型参数的原因,证明了若采用非过零逆重复M序例做激励信号,则仍然可以用相关方法估计其模型参数,并导出了这一模型参数估计的新相关算法。此算法还可用于估计一般的H模型参数。
After introducing a class of even-H models whose output is only related to the square of their input, the reasons why the existing correlation algorithms can not estimate the parameters of this model are analyzed. It is proved that if non-zero-crossing inverse M sequences are used as excitation signals, It is still possible to estimate its model parameters using the correlation method and to derive a new correlation algorithm for this model parameter estimation. This algorithm can also be used to estimate general H-model parameters.