On the use of minimal parametrizations in multivariable output-error identification

Multivariable output-error identification does not constitute, in any way, a straightforward extension of the scalar case. The aim of this paper is twofold: 1) Introduction of a new minimal parametrization for multivariable output error models leading to an easily implementable prediction error identification procedure; 2) Comparison of PEM and errors in-variables approaches based on the dynamic Frisch scheme in the identification of MIMO output error processes.