A Control System for Long-Term Ventilation of the Lungs

A respirator control system based on a variant process model and optimization of system performance is described. The system attempts to minimize the harmful effects of positive pressure ventilation while meeting the ventilatory requirement of the patient. As alveolar pressure is indicative of respiratory dynamics, it has been used as control parameter. Desired alveolar pressure is derived from a fixed parameter RC lung model while actual alveolar pressure is estimated from the variant lung model which is continuously updated through on-line computation of respiratory mechanical parameters. The controller gain is optimally adjusted so as to minimize error index. The system has been simulated on a digital computer and several representative cases of sudden and gradual parameter variation have been studied. It has been shown that in case of changes in the process, the error quickly damps out to zero.