Identification algorithm for MIMO switched output error model in presence of bounded noise

Abdelhak Goudjil (1) , Mathieu Pouliquen (1) , Eric Pigeon (1) , Olivier Gehan (1)

1 LAC - Laboratoire d'automatique de Caen

In this paper, we present an output error identification algorithm for multi-input/multi-output (MIMO) switched linear systems with bounded noise on the output. To our knowledge, no identification algorithm has been proposed for the identification of MIMO-Switched Output Error (MIMO-SOE) model. This problem includes the estimation of the discrete state that indicates the active sub-model at each time and the estimation of the parameters vectors of all sub-models. The proposed algorithm is an adapted Outer Bounding Ellipsoid (OBE) type algorithm suitable for system identification with bounded noise. We show by a numerical example that our algorithm performs well and provides good performance with low computational complexity.

Mots clés: Identification; Linear systems; MIMO systems; Switching systems (control); Switched linear system; MIMO; Output error identification; Bounded noise; Discrete state

Domaines: Automatique / Robotique

Abdelhak Goudjil, Mathieu Pouliquen, Eric Pigeon, Olivier Gehan. Identification algorithm for MIMO switched output error model in presence of bounded noise. 56th IEEE Conference on Decision and Control 2017, Dec 2017, Melbourne, France. pp.5286-5291, ⟨10.1109/CDC.2017.8264441⟩. ⟨hal-01824204⟩