A Robust Interval Estimator--Based Control Design for Anesthesia Delivery Machines

This paper proposes an Interval Estimator–based controller design for anesthesia delivery machines. The Interval Estimator design is based on the positive structural properties of the system obtained from the drug interaction kinematics and dynamics. The state of the estimator is then used to produce a control input, ensuring that despite some class of perturbations or uncertainties, the Bispectral Index (BIS) converges to a region close to the desired value. Moreover, the design does not require linearizing, filtering, or approximating the nonlinear BIS function to produce the regulating feedback. Additionally, a fault detection scheme is proposed to complement the overall design. Finally, some simulation results illustrate the feasibility of the proposed control approach.