摘要:AbstractIn this paper we analyse the use of different models to describe the effects of propofol in the induction of hypnosis on a patient during surgery. In particular, we consider the standard three-compartmental pharmacokinetics/pharmacodynamics Wiener model and a suitable reduction of it. The estimation of the parameters is based on real bispectral index scale surgical data and it is performed by using genetic algorithms. The results obtained for different patients show that the standard model fits the data better than the reduced model, but it is overparameterized and its structure is much more complex to be used for the control design. Conversely, the simplification of the identification procedure obtained with the reduced model, which has only three parameters, is paid by only a slight decrement of the performance.