Title :
Adaptive inverse model control of pressure based ventilation
Author :
Borrello, Michael A.
Author_Institution :
Puritan Bennett Corp., Carlsbad, CA, USA
Abstract :
For the control of critical care ventilators, the problem of tracking a desired pressure trajectory within a patient connecting circuit is solved using indirect adaptive control. This control uses an inverse model of patient lung mechanics and the ventilator connecting circuit. Online estimates of physical parameters are provided using standard recursive least squares with forgetting factor. The indirect method achieves robust performance over a wide range of patient conditions, and also since physical estimates of respiratory parameters are available, they can be applied to diagnostics and/or additional control measures for ventilation. This control method has been applied to NPB 840 ventilator hardware in the loop using the real time simulation software, VisSim. Experimental results on mechanical test lungs show a large improvement in uniform tracking by the adaptive controller over classical PID control methods currently used throughout the ventilator industry
Keywords :
adaptive control; biocontrol; computerised control; digital simulation; health care; least squares approximations; lung; medical computing; pneumodynamics; real-time systems; NPB 840 ventilator hardware; VisSim; adaptive controller; adaptive inverse model control; classical PID control methods; critical care ventilator control; forgetting factor; indirect adaptive control; indirect method; inverse model; mechanical test lungs; online estimates; patient conditions; patient connecting circuit; patient lung mechanics; physical estimates; physical parameters; pressure based ventilation; pressure trajectory tracking; real time simulation software; respiratory parameters; robust performance; standard recursive least squares; ventilator connecting circuit; ventilator industry; Adaptive control; Circuits; Inverse problems; Joining processes; Lungs; Parameter estimation; Pressure control; Programmable control; Trajectory; Ventilation;
Conference_Titel :
American Control Conference, 2001. Proceedings of the 2001
Conference_Location :
Arlington, VA
Print_ISBN :
0-7803-6495-3
DOI :
10.1109/ACC.2001.945900