The Anaerobic Digestion Process Control Using Data Driven Methods

The paper deals with the use of data-driven methods for the synthesis of the anaerobic digestion process control laws, using data gathered from the simulation of a complex model treated as a virtual plant. This approach is used to design local controllers from the gain scheduling strategies. In the paper, two synthesis procedures of the control algorithm are used: a classic data-driven algorithm that is adapted to the particularities of the data used, and a simplified algorithm proposed by the authors, which ensures the interactivity in the synthesis of the control. An illustration of the two synthesis procedures of the control algorithms is made for two operating regimes in which the process has different static and dynamic properties. The important aspect of applying these methods to the anaerobic digestion process control is that the synthesized controller has a non-parametric Bode characteristic of PI-type from which the transfer function of an equivalent PI classical controller is immediately obtained.