On the maximum likelihood method of identification
IBM Journal of Research and Development
Convergence properties of the generalised least squares identitication method
Automatica (Journal of IFAC)
System identification-A survey
Automatica (Journal of IFAC)
On the problem of ambiguities in maximum likelihood identification
Automatica (Journal of IFAC)
Convergence properties of the generalised least squares identitication method
Automatica (Journal of IFAC)
Automatica (Journal of IFAC)
Brief paper: Some properties of the output error method
Automatica (Journal of IFAC)
Necessary and sufficient conditions for uniqueness of the minimum in Prediction Error Identification
Automatica (Journal of IFAC)
Input design as a tool to improve the convergence of PEM
Automatica (Journal of IFAC)
Hi-index | 22.15 |
The maximum likelihood method of identification is a powerful tool for obtaining mathematical models of dynamic processes. To apply this method a loss function has to be minimized. The aim of the paper is an investigation of the local minimum points of this loss function for a common structure of a general form. If the loss function has more than one local minimum point, numerical problems can occur during the minimization. Sufficient conditions are given for the existence of a unique stationary point, which then also gives the desired global minimum. It is also shown by counter-examples that there are systems without peculiarities, which have more than one local minimum point of the loss function.