Title :
Periodic funnel-based control for peak inspiratory pressure
Author :
Anake Pomprapa;Sören Weyer;Steffen Leonhardt;Marian Walter;Berno Misgeld
Author_Institution :
Medical Information Technology, RWTH Aachen University, D-52074, Germany
Abstract :
Intensive care unit patients generally require a mechanical ventilator with the aim to improve gas exchange. Hereby, the quality of artificial ventilation significantly depends on the mechanical ventilator in terms of its stability, accuracy and transient response. In this article, we demonstrate a non-identifier based control algorithm, called funnel-based control of peak inspiratory pressure (PIP). In pressure-controlled ventilation based on our tailor-made mechanical ventilation system, PIP can be periodically regulated to the pre-defined target reference value using an output feedback structure, so that airway pressure can be delivered into the front-end respiratory system. The goal of the control system is to guarantee a fast response with minimal steady state error on a breath-by-breath basis and to suppress the overshooting of the response, in order to avoid alveoli overinflation and possible lung injury. The control performance is assessed with distinctive results from a number of experiments with a plastic lung, based on the hyperbolic and exponential funnel performance boundaries.
Keywords :
"Ventilation","Lungs","Valves","Mathematical model","Transient response","Output feedback"
Conference_Titel :
Decision and Control (CDC), 2015 IEEE 54th Annual Conference on
DOI :
10.1109/CDC.2015.7403100