NEURO-FUZZY CONTROLLER IMPLEMENTATION FOR AN ADAPTIVE CATHODIC PROTECTION ON IRAQ-TURKEY CRUDE OIL PIPELINE

Impressed current cathodic protection is widely used to prevent corrosion of structural steels, especially, those used in a crude oil pipeline. The potential required for cathodic protection depends upon the steel being protected from the environment. Since corrosion rate depends on current flow to or from the steel and the adjacent medium, the corrosion activity of the steel in soil normally increases as soil resistivity decreases. The major constituents are resistivity, chlorides, sulfates, and the acidity (pH) of the soil. Output voltage, depending on desired reference voltage, needs to be controlled and tuned by transformer rectifier (TR) units. Proportional–integral (PI) and proportional–integral–derivative (PID) controllers have been used recently in tuning the output voltage, but overshoot and oscillation occurred in the outputs.

In order to overcome these problems, a new approach based on a neuro-fuzzy controller for tuning the output voltage of TR units is presented in this study. The neuro-fuzzy classification based on a fuzzy logic controller was implemented in a microcontroller. The controller helps to compensate for the variations and the dynamic changes in the TR units on the Iraq-Turkey crude oil pipeline. The experimental results achieved from two stations have shown that the proposed neuro-fuzzy controller is simple, does not require any derivative function, and has no overshoot and oscillation.