Identifiability implies robust identifiability

In identification from a deterministic point of view an algorithm is said to be robustly convergent if the true system is regained when the noise level tends to zero. In this paper we introduce a concept close to this performance measure: robust global identifiability. A model structure, i.e. a smoothly parametrized set of models, is said to be robustly globally identifiable if there exist an identification algorithm such that the true parameters are regained when the noise level tends to zero. We show that global identifiability implies robust global identifiability when the model structure in consideration is a characteristic set of differential polynomials