Study on sulfate reducing bacteria detection using Adaptive Neuro-fuzzy Inference System

The detection of sulfate reducing bacteria (SRB) in a water system is very crucial to prevent the corrosion of iron material in the system. In this regard, a method of using an Adaptive Neuro-fuzzy Inference System (ANFIS) is studied for the modeling and detection of SRB in a medium. A study on ANFIS concept is made to further understand the structure and criteria of the system. The experimental data obtained from data acquisition board are used for training of the ANFIS system. Three parameters (voltage, temperature and humidity) are selected as major factors in determining existence of the bacteria. Two membership functions (trapezoidal and bell-shaped) are used for training the data. The results show that ANFIS with trapezoidal membership function is the best with its average error, 1.66E-07 at epoch 250.