A Comparative Evaluation of Three On-Line Identification Methods for Respiratory Mechanical Model

Three recursive algorithms for on-line identification of a respiratory mechanical model are compared both in deterministic and stochastic environments. These algorithms include: 1) a popular algorithm belonging to the equation error class, 2) an algorithm belonging to the so-called two-stage least-squares (2SLS) schemes, and 3) a more recent algorithm derived from output error configurations. All these algorithms are presented in a unified general formulation.