离散系统参数自适应辨识的最小二乘正交变换方法

An Adaptive ldentification Method with Orthogonal MatrixTransformation Based on Least Square Approximation

基于最小二乘法在消除干扰方面的应用，应用矩阵变换及数值拟合的原理，充分考虑随机信号的影响，给出了一种正交变换的离散差分模型自适应辨识方法，它能自动判别模型的阶，还能体现模型参数随信息更新而实时更新的动态过程。提出了渐近消除记忆的时变离散系统参数自适应辨识方法。

Based on the application of least spure approximation method in eliminatingdisturbance, using the principles of matrix transmation and numerical fitting, considering the in-fluence of random signals, a kind of adaptive identification method for estimating parameters inone-dimensional discrete system is given, by which the rank of the models can be identifiedautomatically, and the dynamic process of the model parameters updating together withinformation can b e reflected. Meanwhile, the adaptive identification method of approximatingmemory-eliminating for estimating parameters in time-varying descrete system is put forward.