BP neural network predictive model for stray current density of a buried metallic pipeline

The purpose of this paper is to analyze and estimate the stray current corrosion hazard of a buried metallic pipeline using a predictive model for stray current density.

A predictive model for stray current density of the buried metallic pipeline was built, using a back propagation (BP) neural network method and experimental data. The accuracy of the model was tested using test samples. The single sensitivity analysis predictive method was used to establish the relationship between stray current density with the soil resistivity. The effects of buried depth and the pipe‐to‐ground voltage offset were researched using this network model.

The feasibility of the BP neural network to forecast stray current effects from the buried metallic pipeline was confirmed.

The paper provides a new method to analyze and estimate the stray current corrosion hazard of buried metallic pipelines.