Assessment of data-driven modeling strategies for water delivery canals

Conferência: CONTROLO’2012 - 16-18 July 2012 - Funchal

The aim of this work is to develop nonlinear dynamical models for the canal system of Núcleo de Hidráulica e Controlo de Canais. The canal is a nonlinear system and thus should be modeled to meet given operational requirements, while capturing all relevant system dynamics, such as the resonance waves created due to the movements of gates, and also contributing to the controller precision. The nonlinear modeling is based on data-driven methods, namely Composite Local Linear Models, Fuzzy Models and Artificial Neural Networks. These models are identified using data collected from the experimental facility, and their performance is assessed based on suitable validation criteria. The modeling results show the effectiveness of these models while capturing all significant dynamics for the canal system.