A simulation model of infant-incubator-feedback system with humidification and temperature control

Abstract

A comprehensive simulation model for the infant-incubator-feedback system was developed in a Matlab/Simulink® environment to investigate all heat exchange relationships, variables and factors that have an influence on the overall thermo-neutrality of the environment. The model was also used to determine the benefits and limitations of using a convectively heated single-walled incubator in nursing preterm infants with very low birth weight < 1000 grams and low gestational age 28 weeks. The infant was modelled as one-lump with two layers; core and skin. The infant shape was approximated to a cylinder. The model incorporated all compartments of the infant-incubator system including core, skin, incubator air space, mattress, incubator walls, air-circulating fan, heating element, added oxygen (for resuscitation purposes), and humidification chamber, which has not previously been considered. The results of the simulation were in terms of the temperature variation over time, of the following parts of the system: core and skin temperatures and incubator air space temperature. Results of the simulation with added humidity showed that the body temperature of a 900 gram infant, with an initial body temperature of 35.5 ºC, did not reach the thermo-neutral range between 36.5-37.5 ºC in two hours, on air mode. Whereas, on skin mode, both core and skin temperature reached to 36.87 ºC and 36.5 ºC in two hours, and thus a thermo-neutral environment was achieved. These outcomes are consistent with clinical empirical reports. The simulation model is a closed-loop system with a PID controller for each mode; air servo controlled and skin servo controlled. The controller parameters were virtually estimated by the Zeigler-Nichols Method as real values were not available. Nevertheless, the overall stability of the whole system has been achieved by applying a step input which was verified by the Root Locus Method.