On the maximum likelihood method of identification
IBM Journal of Research and Development
Brief paper: Test of pole-zero cancellation in estimated models
Automatica (Journal of IFAC)
Tests for determining model order in parameter estimation
Automatica (Journal of IFAC)
The determination of the orders of process-and noise dynamics
Automatica (Journal of IFAC)
Comparison of different methods for identification of industrial processes
Automatica (Journal of IFAC)
On the problem of ambiguities in maximum likelihood identification
Automatica (Journal of IFAC)
Estimation of the order of linear systems
Automatica (Journal of IFAC)
Development of omni-directional correlation functions for nonlinear model validation
Automatica (Journal of IFAC)
Technical communique: On a method for model structure selection in system identification
Automatica (Journal of IFAC)
Special section system identification tutorial: Maximum likelihood and prediction error methods
Automatica (Journal of IFAC)
Brief paper: Experiment design for maximum-power model validation
Automatica (Journal of IFAC)
Hi-index | 22.15 |
A new solution is presented to the problem of validating optimally a given dynamic model against given long-sample observations. If the model can be parametrized and cast into a general innovations structure, i.e. if expressions for the one-step predictor and the prediction error covariances are available, a test can be constructed that has asymptotic maximum discriminating power, for the least favourable case that the difference to be detected between model and observed system is small. A class of alternative models must be specified, but, unlike in other optimal tests, it is not required also to fit a best model within this class. Since the alternative class may include models more complicated than that to be validated, the test can be used for recursive determination of structure and order. For linear transfer-function or polynomial-operator models the asymptotic maximum-power test does not require much more computing, and sometimes less, than the conventional tests of auto- and cross-correlation. Generally, the latter tests are less efficient, even for linear models, if these is some a priori knowledge about the structure. A simple example demonstrates that there are realistic cases where the asymptotic maximum-power test may be considerably better.