【摘要】 对于一个带未知噪声的多传感器系统,当各种噪声方差和噪声之间的协方差都不为0时,提出了一种基于自回归滑动平均(ARMA)模型的多段辨识算法.使用该算法可以得到对系统未知参数和未知噪声方差和协方差的无偏估计,并且证明了估计的收敛性.并通过一个Matlab仿真例子对算法的有效性做出了证明.更多还原

【Abstract】 For the multi-sensor system with unknown noises,when all kinds of the noise variance and covariance are not 0,based on the Autoregressive Moving Average(ARMA)innovation models,a multistage identification algorithm is presented.Using this algorithm can get unbiased estimations of the unknown parameters and the unknown noise variance and covariance and the convergence of the estimations is proved.A Matlab simulation example shows the effectiveness of the algorithm.更多还原