Oceanic Atmospheric Environment Equipment Metal Component Corrosion Rate Prediction Research

Objective: To address the issue of predicting the corrosion rate of metal components in the marine atmosphere environment, a model combining the improved dung beetle optimization algorithm (CDBO) and the kernel extreme learning machine (KELM) was established to improve the prediction accuracy of metal corrosion rates. Method: By introducing chaos mapping strategies, adaptive weight golden sine strategies, etc., to overcome the problems of the dung beetle algorithm being prone to local optimum and slow convergence speed, the parameters in KELM were optimized using CDBO to establish the optimization model. Finally, 60 sets of corrosion rate data of metal structural components were used for verification. Result: The prediction results of the CDBO-KELM model were closer to the true values than other models, with better stability, and the MAE, MAPE, and MSE indicators were all superior to the comparison models. Conclusion: This model can predict the corrosion rate of equipment structural components more accurately and reliably, providing a theoretical reference for equipment corrosion control and management in the marine atmosphere environment.