CHAPTER 3
DATA SETS AND MODELS FOR THE DATA
The data used in this thesis belongs to the biomedical field, but this application is
only one of many that can come to mind when thinking about real-world signals. Most of
the simulations results are obtained on a particular time series presented below.
3.1 Real World Non-Stationary Data
3.1.1 Description of the Data
Ventilators are used for patients who's respiratory muscles are too weak to breathe
consistently on their own. A ventilator's goal is to optimize the patient's work of
breathing (i.e., the ventilator has to be set so that the patient is producing just the right
amount of effort to take a breath). In order to optimize the parameter settings of the
ventilator and the comfort of the patient, a physician needs to have a good model of the
patient's lung and to be able to monitor the dynamics of the respiratory system.
The simplest model is called the One Compartment Model, because it does not
make a differentiation between the two lungs. This model is defined by (3.1), where P is
the driving pressure in cmH20, R is the resistance to flow (or variation in volume) in
cmH20.Ll .sec V is the airflow in L.sec1, E is the elastance in cmH20.L1 and V the
volume in L.
P(t) = R V(t) + EV(t) (3.1)