Consistency of in-situ brass corrosion in HCl solution image fluctuations and electrochemical potential noise revealed through NARX neural network

• DOI: 10.1016/j.jmrt.2024.01.259
• Publication Date: 2024-02-02T11:59:57Z
• Authors: Zhiqin Wu; Haofeng Zhang; Kaixuan Feng; Hong Yan; Honggun Song; Chao Luo; Zhi Hu
• Keywords: Electrochemical noise; Brass; Corrosion monitoring

Micro-scale corrosion imaging in conjunction with electrochemical potential noise (EPN) provides a promising, comprehensive monitoring technique that can precisely visualise specific aspects of activity. In this work, rapid, economical, and real-time through self-designed device, captured simultaneously the micrographs, was employed to analyse activity H57 brass an HCl solution using gray-level co-occurrence matrix (GLCM). Conducted time-frequency domain, continuous wavelet transform (CWT) analysis potentials image features shows EPN fluctuations corresponding feature noises. Similarly, consistency them in-situ Brass solution, as revealed nonlinear autoregressive exogenous (NARX) neural network. The results indicate extracted by GLCM quantify changes images correspond behaviour brass. Notably, anomalous peaks noises accompany or lag behind EPN. Leveraging these features, optimized NARX model demonstrates high predictive capability for These findings are expected guide future techniques even enable prediction via micron-scale images.