摘要
冲激响应反应了线性系统的内部动态特性。该文详细介绍了系统冲激响应辨识的时域互相关技术。它用白噪声对系统进行驱动,将离散时间线性系统转化为时间序列的ARMA模型,并将基于MA模型的系统辨识问题转化为一个优化问题,以最小二乘标准求解。求解过程利用采样样本估计相关函数,构造对称的Toeplitz矩阵方程,再用迭代的Levinson-Durbin算法解该矩阵方程。最后,文章给出了一个线性系统的测量实例。实验结果表明,该方法能准确的估计线性系统的冲激响应。
The impulse response reflects the internal dynamic characteristics of a linear system.The time domain cross-correlation technique for identifying the impulse response of a linear system was introduced.White noise is used to stimulate the unknown system,which,in this case,is equivalent to a time series ARMA model.MA model is suitable to describe the unknown system,because the coefficients of the model equal to the values of the discrete impulse response sequence.An optimal problem is constructed to solve the identification problem and the Least Square criterion is used,yielding a symmetric Toeplitz matrix equation,which is then solved with the recursive Levinson-Durbin algorithm.In the end,an example of the measurement is provided.The result shows the time domain cross-correlation technique can be used to estimate the impulse response of a linear system.
出处
《计算机与数字工程》
2012年第2期1-3,34,共4页
Computer & Digital Engineering
基金
国家自然科学基金项目(编号:60571011
30327001)资助
关键词
白噪声
冲激响应
互相关技术
系统辨识
ARMA模型
white noise
impulse response
cross-correlation technique
system identification
AMRA model
