Control applications using computational intelligence methodologies

The ambiguity, robustness and performance of systems becomes increasingly evident as the complexity and size increase in industrial processes. Under these circumstances, conventional control techniques such as PID control, nonlinear feedback control, LQG control and Hinfin control have several disadvantages because they are hard or inflexible and cannot handle qualitative knowledge. This may explain the dominant role of emerging "intelligent control, using soft computing" as an interesting alternative. This means that increasing complexity requires an increase in intelligence in order to provide robustness and adaptability. Intelligent techniques such as fuzzy systems (FS), neural networks (NN) and genetic algorithms (GA) can be combined using the resources of such techniques complementary. This paper is a report of a research work in intelligent control of the GISIA group (Grupo de Investigation en Sistemas Inteligentes y Aplicaciones) at Escuela Politecnica Nacional