Identification of an Active Suspension in the Presence of Narrow Band Disturbances

Dan Stefanoiu (1) , Ioan Doré Landau (2, 3) , Mihai Cornoiu (1)

1 Department of Computer Science [Bucharest]
2 CENER - Centro Nacional de Energías Renovables - Fundación CENER-CIEMAT
3 GIPSA-SLR - GIPSA - Systèmes linéaires et robustesse

This paper aims to describe a comparative study regarding several identification models and control solutions for an active suspension system, while corrupted by narrow band disturbances. The tested identification models are linear and belong to ARMAX class. Several identification methods were compared in terms of disturbances rejection, both in open and closed loop cases. Beside the Least Squares class of methods, Minimum Prediction Error and Minimum Output Error methods have been investigated as well. The subsequent controller has been designed for the best found models and is either of RST or Q-parameterized class. After several simulations with real data, one has found that models estimated by means of Generalized Closed Loop Output Error (GCLOE) Method are seemingly the best ones.

Mots clés: "Active vibration control"; " system identification"; "digital control"; digital control; system identification; Active vibration control

Domaines: Automatique

Dan Stefanoiu, Ioan Doré Landau, Mihai Cornoiu. Identification of an Active Suspension in the Presence of Narrow Band Disturbances. MCPL 2010- 5th IFAC Conference on Management and Control of Production and Logistics, Sep 2010, Coimbra, Portugal. paper 63. ⟨hal-00712649⟩