Performance Evaluation of ILC Controller for Anesthesia Process

Abstract

An adequate adjust of propofol is crucial to maintain the depth of hypnosis in anesthesia. It is a tedious task of anesthesiologist during surgery. The less amount of propofol infusion leads to insufficient anesthesia and higher amount stands dangerous for the patient. Here, iterative learning controller (ILC) is designed to control depth of hypnosis in anesthesia with BIS as feedback signal. BIS is calculated from EEG signal. In this, automatic propofol infusion system, the patient model, is developed based on pharmacokinetic and pharmocodynamic (PK/PD) model in MATLAB/SIMULINK. The anesthesia process is approximated to transfer function model of first-order time delay. The conventional Internal Model Control-based PID controller (IMC PID) and Ziegler Nichols PID (ZN PID) controller are used to compare the efficiency of proposed ILC controller. The performance results in terms of quality indices (overshoot, settling time) and error indices (Integral Square Error (ISE) and Integral Absolute Error (IAE)) show that the ILC controller is superior than conventional controllers.