An Adaptive Gpc Approach to Low-flow Anaesthesia

An adaptive method is developed to control the end-tidal isoflurane concentration using a low-flow method. The isoflurane concentration is calculated based on measurements of the end-tidal isoflurane concentration and mean arterial pressure (MAP) using a modelbased multivariable GPC algorithm. The control design allows explicit handling of input and output constrains, accommodates model uncertainty and takes into account possible MAP disturbances caused by surgical stimulation and blood loss. The proposed approach was tested on a complex nonlinear compartmental model of the anaesthesia system. The simulations demonstrate that the proposed control approach provides fast response, accurate set-point tracking and has good disturbance rejection behaviour. The advantages and drawbacks of low-flow over high flow anaesthesia techniques were evaluated using controlrelated performance criteria. Copyright © 2005 IFAC