Design of Observer for Cardiovascular Anomalies Detection: an LMI approach

Abstract

In this work, we consider the observer design problem for fault detection in the human cardiovascular system. Our idea of designing an observer for singular systems is based on separating the static equations from the dynamic equations, which simplifies the task of designing the observer by removing the spread of the singularity in the calculus. The objective is to build an observer that simultaneously estimates the states of the system and allows to conclude on the occurrence of faults. By using Lyapunov Method, the convergence of the error of estimation was analyzed and the gain matrix of the observer is got by solving a single linear matrix inequality (LMI). Finally, simulations are carried out and show that the observer efficiently estimates the unknown states and allows us to detect cardiovascular anomalies that may arise from variations in physiological parameters.