双率Hammerstein系统的极大似然递推辨识

Maximum Likelihood Identification of Dual-rate Hammerstein Systems

针对一类双率Hammerstien系统的参数辨识问题,基于辅助模型辨识思想,利用极大似然原理和递推辨识技术,提出一种极大似然递推辨识算法。主要方法是针对模型中的未知输出构造一个辅助模型,用辅助模型的输出预测未知输出。该方法可以直接基于双率输入输出数据进行参数辨识。仿真实验表明,所提出的算法能有效地辨识双率Hammerstein系统,最终误差趋于1%左右。

Aiming at a kind of dual-rate Hammerstein system,based on the auxiliary model identification idea,using the maximum likelihood principle and recursive identification technology,this paper proposes a maximum likelihood recursive least squares algorithm.The main method is to construct an auxiliary model for the unknown output in the model,and use the output of the auxiliary model to predict the unknown output.This method can directly identify parameters based on the dual-rate input and output data.The simulation experiments show that the proposed algorithm can effectively identify the dual-rate Hammerstein system,and the final error tends to be about 1%.