作者: R. H. Shumway , D. E. Olsen , L. J. Levy
DOI: 10.1080/03610928108828137
关键词:
摘要: Kalman filtering techniques are widely used by engineers to recursively estimate random signal parameters which essentially coefficients in a large-scale time series regression model. These Bayesian estimators depend on the values assumed for mean and covariance associated with initial state of signal. This paper considers likelihood approach estimation tests hypotheses involving critical means covariances. A computationally simple convergent iterative algorithm is generate only standard filter outputs at each successive stage. Conditions given under maximum consistent asymptotically normal. The procedure illustrated using typical data set 10-dimensional vectors.