Application of Adaptive Neuro-Fuzzy Inference System Approach for Modelling and Prediction Corrosion Penetration Rate in CrudeOil Pipelines Based on NORSOK Simulation Software

Abstract

This study aims to study the effect of several relevant parameters, namely temperature, pressure, flow rate, and pH, on the Corrosion Penetration Rate (CPR) of the process of transporting crude oil through pipelines through simulation. NORSOK software was used to simulate the experiments and calculate the corrosion penetration rate for each experiment. Experiments were designed based on central Compound Experimental Design (CCD) using Minitab 17 software. Adaptive Neuro-Fuzzy Inference System (ANFIS), were employed to predict the corrosion penetration rate. The predictions obtained from this method was then compared to the actual values acquired through simulation. To evaluate the accuracy of the predictions, the Mean Absolute Percentage Error (MAPE) was calculated. The mean absolute percentage of error for prediction models using ANFIS is 0.0%.