Search

The development of the Artificial Pancreas, which consists of the implementation of a continuous insulin infusion pump and sensors for blood glucose monitoring, is a priority for the treatment of subjects with type 1 diabetes. To achieve a complete automation of this process several control strategies have been proposed with the aim of maintaining blood glucose levels in the normoglycemic range. Among the developed strategies, those based on adaptive control stand out, to counteract the effect of physiological variations that are inherent to the patient. However, these strategies require full state measurement, which is not an easy task because of sensors cost and the inherent difficulty in direct measurement of physiological variables. Therefore, in this work, an adaptive control strategy with guaranteed stability is used for an insulin control model, and a nonlinear sliding mode observer is designed to estimate the state variables that the controller requires. With the complete control-observer system, satisfactory results are obtained by compensating for meal intake and errors between the model and subject parameters. [ABSTRACT FROM AUTHOR]

The design and implementation of a predictive/adaptive control system is presented, using neural engineering techniques to control a non-linear MIMO system in order to control, at a later stage, the temperature and level in a non-linear conical plant. Preliminarily, conventional control structures were tested, which gave rise to the need to test intelligent control structures that allow the control objectives to be met more effectively. The process begins with the experimentation of different neuronal control structures, and then escalates to a predictive/adaptive neuronal control system. The results achieved at the simulation level, testing the proposed design on mathematical models of non-linear MIMO systems, were satisfactory and met the control objectives established, therefore, in the next stage of the project, the experimentation is estimated in the real plant under study. [ABSTRACT FROM AUTHOR]