Automatic control of VFA in anaerobic digestion using a fuzzy logic based law

A control law based on fuzzy logic features was developed and validated for an anaerobic wastewater treatment (AWT) process. The controlled variable was the concentration of Volatile Fatty Acids (VFA) in the anaerobic reactor and the manipulated variable was the input flow rate. In order to use it as the input in a Mamdani-inference fuzzy set, the controlled variable was treated using an algorithm of interpolation, extrapolation and filtering. The treatment of VFA values attempted to anticipate the behaviour of the variable and to avoid the inherent delay of the response, associated to the time constant of the system. Furthermore, the derivative of the controlled variable was used as a second input in the Mamdani-inference fuzzy set as an accelerator or decelerator of the control action. The control law was applied to a 0.948 m of useful volume fixed-bed upflow anaerobic reactor treating raw and diluted (1:2) industrial distillery vinasses. The validation was performed establishing different transient states between different set points in the range of 0.8 and 1.8 g VFA/l and different levels of dilution of the influent (e.g. input Chemical Oxygen Demand-COD concentration). The control law probed to be reliable supplying an adequate control action in terms of amplitude and velocity to achieve the desired set point for different types of perturbation and control purposes.