Novel method for prediction of corrosion current density of gas pipeline steel under stray current interference based on hybrid LWQPSO-NN model

Stray current corrosion poses threaten to gas pipelines, and corresponding test method provides effective preventions to limit the damage. Three-electrode system has limitations in terms of convenience and economy in the practical application. Thus, this work aims to develop a novel non-destructive testing method of corrosion current density in the presence of stray current, in which a network-based model is proposed to realize highaccuracy measurement. An interdisciplinary approach is presented by combining electrochemical laboratory measurements with data-driven technology to predict corrosion current density. An innovative algorithm combining neural network with Levy Flight Weighted Quantum Particle Swarm Optimization is proposed to predict corrosion current density through non-destructive input factors. Results demonstrate that proposed approach improves algorithm performance significantly, including mean accuracy rate (AR) and stability. The best mean AR of corrosion current density is 94.11 %, enhancing the performance by 7.35 % to 18.16 % in comparison to other network-based algorithm.