FUZZY MODELING AND IDENTIFICATION APPLIED TO A BENCHMARK PROBLEM: pH CONTROL

This work presents an approximation to the modeling and identification tasks using Fuzzy Inference Systems (FIS) applied to the Benchmark Problem: pH neutralization control in industrial processes. The statement of this problem is taken from a paper that postulates the waste water pH neutralization process as a benchmark approach. The proposed model uses the Takagi-Sugeno FIS type, aiming at future applications which may arise from model-based control strategies requiring a good predictive level. The final FIS used for modeling resulted from various criteria applied in obtaining the model’s number of clusters, number and type of inputs and typical Takagi-Sugeno parameters. The FIS here presented performs better than other models for similar processes. Besides, its features were intended to reproduce as near as possible the operative conditions of similar industrial plants.