Neuro-fuzzy Techniques Used for Steady State Modeling of pH Neutralization Process

This paper presents an artificial intelligence approach on modeling highly nonlinear processes with application to pH neutralization process. The hybrid neuro-fuzzy technique is applied to model the titration curve from a weak acid with a strong base. This titration curve has three major parts that describe the system behavior: above the equivalence point, near equivalence point and under equivalence point. Process gain has an important variation due to its high nonlinearity. The ANFIS method provides a good solution in modeling this nonlinear titration curve. The resulted model can be used to control the pH neutralization process. The results are promising and can be further developed for the other-input output channels of the pH neutralization process (acid volume - pH, acid concentration - pH, base concentration -pH).