Modelling of dynamic cerebral pressure autoregulation using sequential genetic algorithm

Abstract

Accurate modelling is desirable for analysis and clinical studies of physiological systems. The present work provides methodology for fully automated sequential genetic algorithm (SGA) for auto regressive exogenous (ARX) modelling. The SGA has been implemented to determine proper model structure and thereafter model parameters. The proposed algorithm has been tested on known ARX model and sunspot data modelling problem. Finally, SGA has been applied to model the dynamic cerebral autoregulation (CA) system. The results are promising and models obtained using SGA are better as compared to standard least square (LS) algorithms and can be reliably applied to model physiological system. © 2010 Inderscience Enterprises Ltd.