In this paper a software sensor for estimating the respiration rate and the nonlinear oxygen transfer functionKLa is presented. The respiration rate and the oxygen transfer function were estimated from measurements of the dissolved oxygen concentration (DO) and airflow rate by a Kalman filter. In particular, a filtering procedure was applied for the case when the DO sensor dynamic cannot be neglected. In the estimation scheme the time varying respiration rate was modelled by a filtered random walk model, and the nonlinearKLa function was modelled with an exponential model. A numerical study illustrated the advantage of the method. Also, real data were applied to the software sensor with promising results.

This content is only available as a PDF.