Open AccessPrediction of Corrosion Rate in Submarine Multiphase Flow Pipeline Based on PSO-SVM Model
Author(s) -
Panfeng Wang,
Qing Quan
Publication year - 2019
Publication title -
iop conference series. materials science and engineering
Language(s) - English
Resource type - Journals
eISSN - 1757-899X
pISSN - 1757-8981
DOI - 10.1088/1757-899x/688/4/044015
Subject(s) - support vector machine , corrosion , pipeline (software) , pipeline transport , submarine pipeline , multiphase flow , volumetric flow rate , seabed , submarine , flow (mathematics) , computer science , petroleum engineering , marine engineering , engineering , artificial intelligence , materials science , geotechnical engineering , geology , metallurgy , mechanics , mechanical engineering , oceanography , physics
In view of the internal corrosion rate of submarine multiphase flow pipelines, this paper analyzes the related factors affecting the corrosion rate of this type of pipeline, and introduces PSO algorithm and SVM algorithm respectively. Based on the PSO-SVM combination model, the 44 groups of data was used to study the influencing factors and corrosion rate, meanwhile the 10 groups of data was used to predict. The predictions are compared with the GA-SVM model, the LS-SVM model and the CV-SVM model to verify the advancement and feasibility of the proposed method. The results show that the temperature has a relatively large influence on the corrosion rate of the multiphase flow pipeline in the seabed. The influence of pressure on the corrosion rate of the multiphase flow pipeline in the seabed is relatively small. The PSO-SVM combined model is used in the submarine multiphase flow pipeline. The e error of corrosion rate prediction is only 1.848% on average, and the model training time is only 3.17s, both of which are smaller than other models. The research proves that the PSO-SVM combination model has strong advancement and feasibility for the prediction of the internal corrosion rate of submarine multiphase flow pipelines.