A comparative study of three advanced controllers for the regulation of hypnosis

The anesthetic state is a dynamic combination of hypnosis, analgesia and neuromuscular blockade that is maintained by infusing a cocktail of drugs. This work focuses on controlling the hypnosis during a surgical procedure by automatic regulation of isoflurane and employs Bispectral Index (BIS) as the primary controlled variable. A seventh-order nonlinear pharmacokinetic–pharmacodynamic representation has been used for the hypnosis dynamics of patients. This study uses a model predictive control structure for the regulation of BIS. Performance of this controller has been tested for a range of patients and compared with two previously employed control strategies (cascade internal model controller and cascade controller with modeling error compensation). Performance of the three controllers has also been studied for a step change in BIS, measured disturbances and noise in the measured variables. Numerical simulations show that the model predictive controller performed better than the other two controllers.