Adaptive Fuzzy Generalized Predictive Control of pH in Tubular Photobioreactors on Microalgae Plant

Culture conditions in a photobioreactor on a microalgae plant are essential for production rates, and the pH value is one of the critical variables for a harmonizing process. However, the biological nature of microalgae growth represents a complex reaction that is arduous to model and control. This work proposes an approach of the application of adaptive identification and predictive control to regulate the pH in a photobioreactor in a microalgae production system. The proposed approach is composed of two main stages: identification and control. The identification is performed by a Takagi-Sugeno (T-S) fuzzy model, which will be learned with an offline strategy, by Fuzzy c-means, for the antecedent part, and by the Least Squares Method (LSM), for the consequent part. Then, the model is updated with an online method, the Recursive Least Squares with Adaptive Directional Forgetting (RLS-ADF) factor algorithm. The control scheme is based on a Generalized Predictive Control approach, which is a Model Predictive Controller (MPC), with the adaptive T-S fuzzy model designed in the identification stage. In this way, the model parameters from GPC are online adapted by RLS-ADF. To validate the control structure, the proposed approach was tested by using an estimated model from real data of a microalgae production process.