Interval Observer-Based Model Predictive Control in the Propofol Delivery Process

Due to the anesthesia delivery-related consequences during general surgery, it is desirable to regulate hypnosis during the surgery through an anesthesia infusion closed-loop system. In this paper, a closed-loop scheme is based on the patient’s kinetic and dynamic drug model, where the Bispectral index is the available output. In order to generate the feedback and recover the necessary states, an interval observer is designed based on the system’s positivity and cooperativity properties. Then, a Model Predictive Control (MPC) is implemented to ensure the correct propofol delivery using the estimated states and output. Finally, the proposed interval observer and MPC performances are validated by simulations in three silico patients.