Discrete-time Neural Observer for HIV infection dynamics

This paper presents a discrete-time neural observer for nonlinear systems, whose mathematical model is assumed to be unknown. The observer is based on a recurrent high order neural network (RHONN), which is trained on-line with an extended Kalman filter (EKF)-based algorithm. The respective stability analysis based on the Lyapunov approach is included. The neural observer is tested by application to an immunological interaction model for HIV. The observer estimates the non-measured number of infected CD4+T cells in the blood torrent, the measured number of non-infected CD4+T cells and the measured concentration of viral load. The observer performance is illustrated via simulations.