Artificial Neural Network Based Control of a Pressure Controlled Artificial Ventilator

Artificial ventilation involves a mechanical ventilator that manages the patient’s respiratory needs by increasing and lowering pressure during inhalation and exhalation. A well-designed ventilator control system ensures accurate pressure maintenance throughout the respiratory cycle. This process, known as patient-ventilator synchrony, minimizes discomfort and prevents complications associated with mechanical ventilation. In this paper, with the aim of enhancing the dynamic stability of a volume-controlled ventilation system, an artificial neural network (ANN) based controller is purposed. The controller employs a well-known multilayer perceptron (MLP) which allows for improved adaptive control and generalize from data, leading to more accurate and responsive ventilation support suited for the patient’s needs. This aims to provide better synchronization between the ventilator and the patient’s breathing patterns, ultimately enhancing patient comfort and safety.