KAUST DepartmentComputer, Electrical and Mathematical Sciences and Engineering (CEMSE) Division
Applied Mathematics and Computational Science Program
MetadataShow full item record
AbstractThis paper discusses the possibility of using observer-based approaches for cardiovascular anomalies detection and isolation. We consider a lumped parameter model of the cardiovascular system that can be written in a form of nonlinear state-space representation. We show that residuals that are sensitive to variations in some cardiovascular parameters and to abnormal opening and closure of the valves, can be generated. Since the whole state is not easily available for measurement, we propose to associate the residual generator to a robust extended kalman filter. Numerical results performed on synthetic data are provided.
Conference/Event name13th European Control Conference, ECC 2014