Diseño y simulación de un controlador por modos deslizantes adaptativo; basado en multi-modelo difuso de Takagi-Sugeno, junto con predictor de Smith aplicado a un proceso de neutralización de pH.

Abstract

This work presents the design and simulation of an adaptive sliding mode controller based on a Takagi-Sugeno fuzzy model, alongside a Smith predictor, applied to a pH neutralization reactor, a non-lineal and dominant and variable dead-time process. First order plus dead-time models are obtained around various operating points of the plant, these models are used in a Takagi-Sugeno fuzzy inference system to represent the non-linear dynamic of the pH neutralization reactor. This is used as a Smith predictor to solve the problems caused due to the system dead-time. Based on the Smith predictor signal, a sliding mode controller is designed, which adapts through a gain-scheduled scheme using the fuzzy model of the system. To test the functioning of the system, regulation and tracking tests to step changes in the reference are performed. The designed scheme is also quantitively compared with a PID controller and a SMC based on first order plus dead time models, using ISE and TVu performance indices.