论文部分内容阅读
在高阶卡尔曼滤波应用中,分析人员常常对系统观测能力的性质了解甚微。例如,在有些情况下,滤波器虽能精确地估计某种线性联合的状态变量,但却不能明显地看出误差协方差矩阵。本文指出,适当地归一化后的误差协方差矩阵的特征值和特征向量可以提供关于系统观测能力的有用信息。
In high-order Kalman filter applications, analysts often have little idea of the nature of the system’s observational capabilities. For example, in some cases, the filter can accurately estimate a linear combination of state variables, but can not clearly see the error covariance matrix. This paper points out that the properly normalized eigenvalues and eigenvectors of the error covariance matrix can provide useful information about the observability of the system.