Fuzzy Modeling of a Nonlinear Continuous Fermentation Bioreactor for Ethanol Production

Ethanol production from glucose was analyzed implementing fuzzy techniques to investigate and model the nonlinear behavior of the process. The model simulated involves the basic equations of a chemostat including the dependence of kinetics parameters on temperature and mass transfer of oxygen. For the identification of nonlinearities. Steady-state values for inlet temperature, the flow of refrigerant and the initial concentration of substrate are 25 ° C, 18 L / h and 60 g / L glucose, respectively. The implementation of the Fuzzy Inference System (FIS) was based on a Mamdani inference engine, and also that the defuzzification method implemented was centroid and the weight of the trapeziums describing the output from 15 to 30 g/L affects considerably. After evaluating and comparing the results of the simulation with FIS results, and calculating the correlation between Predicted and Expected values, it is concluded that correlation factor for the entire simulation and FIS was around 0.9. The lack of fitness is evaluated and analyzed after the comparison of results.