| LUO Zheng-shan,QIN Yue,ZHANG Xin-sheng,BI Ao-rui.Prediction of External Corrosion Rate of Marine Pipelines Based on LASSO-WOA-LSSVM[J],50(5):245-252 |
| Prediction of External Corrosion Rate of Marine Pipelines Based on LASSO-WOA-LSSVM |
| Received:March 28, 2020 Revised:June 04, 2020 |
| View Full Text View/Add Comment Download reader |
| DOI：10.16490/j.cnki.issn.1001-3660.2021.05.027 |
| KeyWord:marine pipeline external corrosion rate prediction model LASSO regression method whale optimization algorithm (WOA) least squares support vector machine (LSSVM) |
| Author | Institution |
| LUO Zheng-shan |
School of Management, Xi’an University of Architecture & Technology, Xi’an , China |
| QIN Yue |
School of Management, Xi’an University of Architecture & Technology, Xi’an , China |
| ZHANG Xin-sheng |
School of Management, Xi’an University of Architecture & Technology, Xi’an , China |
| BI Ao-rui |
School of Management, Xi’an University of Architecture & Technology, Xi’an , China |
|
| Hits: |
| Download times: |
| Abstract: |
| The paper aims to construct a prediction model for the external corrosion rate of marine pipelines and improve the accuracy of prediction for the corrosion rate of offshore pipelines, the least square support vector machine (LSSVM) corrosion rate prediction model based on LASSO (LASSO) regression algorithm and whale optimization algorithm (WOA) was established. The index was screened by LASSO regression method to extract the main influencing factors of marine pipeline corrosion. The least square support vector machine algorithm is used to establish a prediction model for the corrosion rate of submarine pipelines, and the whale optimization algorithm is used to optimize the model parameters to avoid the influence of parameter values on the regression performance of the model. Based on the experiment of real sea hanging film, the simulation is carried out through MATLAB, and the prediction results of the model are analyzed and verified, and the prediction results are compared with other models. The main factors obtained by LASSO regression algorithm are temperature, dissolved oxygen content and pH value. The prediction results of the adopted WOA-LSSVM model are in good agreement with the actual values. The average relative error is 2.23%, the root mean square error (RMSE) is 0.3248, and the coefficient of determination R2 reaches 0.9708, which are better than the other two models. The prediction model of the least square support vector machine based on LASSO regression and whale optimization algorithm has better generalization ability and prediction accuracy, which provides a new idea for the research of submarine pipeline corrosion, and also provides a reference for the structural safety and risk prevention of marine oil and gas transportation system. |
| Close |
|
|
|