Petrochemical equipment corrosion prediction based on BP artificial neural network

In order to avoid the accident in refinery due to petrochemical equipment corrosion, this paper researched a BP neural network corrosion prediction model on petrochemical equipment. This model can better express the relationship of the corrosive medium factors (PH, CL-, H₂S, NH₃N) and the corrosion products(Fe²⁺ and Fe³⁺). Simulation results show that the predicted data are close to the monitored data, and the maximum relative prediction error is about 10%, so this model can be used to predict corrosion on petrochemical equipment.