对于在噪声环境下的反卷积系统的多传感器信息融合辨识算法

An Information Fusion Identification Algorithm for the Deconvolution System under Noisy Environment

对于一个在噪声环境下的反卷积系统,当噪声为未知白噪声时,提出了一种多传感器信息融合辨识算法。该算法的核心是依次使用递推增广最小二乘法、相关函数法和Gevers-Wouters算法。使用该算法可以得到对系统未知参数和未知噪声的局部和融合估计,并且证明了辨识的收敛性。用Matlab软件对一个例子进行仿真,从而对算法的有效性做了说明。

For a deconvolution system with unknown white noises,when the model parameters and noises statistics are unknown,a multi- sensor information fusion multi- stage identification algorithm is presented. This algorithm is constituted by recursive extended least squares method,correlated function method and Gevers- Wouters algorithm. Using this algorithm can get the local and fused estimations of the unknown parameters and unknown noises statistics. The convergence of the estimations is proved. An example shows the effectiveness of the algorithm.