Corrosion Monitoring Based on Recurrence Quantification Analysis of Electrochemical Noise and Machine Learning Methods

Abstract

Although electrochemical noise (EN) has been studied for decades, the optimal approach for the analysis of EN data remains uncertain. This research innovatively combined the use of recurrence quantification analysis of electrochemical noise data and machine learning methods to develop models for corrosion monitoring and corrosion type identification. Case studies demonstrate that the proposed methodologies are potentially feasible for the development of online corrosion monitoring programs.