Feedback control of HIV antiviral therapy with long measurement time

In this presentation we apply a receding horizon observer to an HIV feedback control problem in order to derive optimal treatments of HIV pro- gression and/or optimal structured treatment interruptions (STI) in antiviral therapy that include drug-free periods of immune-mediated control of HIV. We use a nonlinear differential equation model which has been well-validated with clinical patient data and shown to have reliable predictive capabilities. The basic feedback control problem utilizes a tracking formulation (desired states for viral load and immune effector calculated from a healthy steady state for the model are tracked). Here we assume only very realistic observerables of CD4+ T cell count and viral load. Moreover, we use a second deterministic optimal tracking problem for state estimation as opposed to stochastic filtering approaches. In both of the optimization problems (for optimal feedback control and for state estimator) conjugate gradient algorithms are employed.