Prediction of Corrosion Rate of Submarine Oil and Gas Pipelines Based on IA-SVM Model

Abstract

In view of the corrosion problem of submarine oil and gas pipelines, this paper proposes a predictive model based on support vector machine (SVM), according to the related factors affecting the corrosion of submarine oil and gas pipelines, and uses the immune algorithm based on the problem of relatively low corrosion rate and influencing factors. (IA) prefers the penalty parameters C and g of the SVM. The IA-SVM model is combined to form the IA-SVM model, and the number of hidden nodes and the kernel function of the SVM are optimized based on the absolute error. Finally, the model is verified according to the actual corrosion rate of the submarine pipeline in a certain sea area of China, and with the PSO-SVM, the GA-SVM and LS-SVM models are used to compare the prediction errors to verify the feasibility and advancement of the IA-SVM model. The research shows that the preferred results of IA for SVM penalty parameters C and ϵ are 43, 6213 and 0.0483, the preferred result of SVM hidden layer nodes is 260, and the kernel function preferred result is Sigmoidal function. At this time, the predicted mean absolute error and root mean square error of the combined model are 1.45% and 0.0159165, respectively, the error of the model is smaller than other prediction models. The research results show that the prediction error of the corrosion rate of submarine oil and gas pipelines based on IA-SVM model is relatively small, and the data training time is short, which can be used to predict the corrosion rate of submarine oil and gas pipelines.