Open Access

ARTICLE

Residual Strength Prediction of Corroded Pipelines Based on Sparrow Search Algorithm-Optimized Kernel Extreme Learning Machine

Zixuan Zong¹, Tingting Long¹, Huaqing Dong^1,*, Guoqiang Huang¹, Xiao Meng¹, Mohammadamin Azimi²

1 Sinopec Sichuan-East Gas Transmission Branch, Wuhan, China
2 GeoEngineers Inc., San Diego, CA, USA

* Corresponding Author: Huaqing Dong. Email:

(This article belongs to the Special Issue: Greening the Pipes: Achieving Sustainability in Pipeline Engineering)

Structural Durability & Health Monitoring 2026, 20(3), 21 https://doi.org/10.32604/sdhm.2026.077489

Received 10 December 2025; Accepted 28 January 2026; Issue published 18 May 2026

Abstract

This paper proposes a novel approach for predicting the residual strength of corroded pipelines by combining the Kernel Extreme Learning Machine (KELM) with Sparrow Search Algorithm (SSA) optimization. The proposed SSA-KELM model addresses the limitations of traditional evaluation methods and single machine learning models in residual strength prediction. A dataset comprising 80 samples from burst tests and finite element simulations was used to validate the model. Results demonstrate that the SSA-KELM model achieves superior prediction accuracy with a maximum relative error of 13.54% and minimum relative error of 0.20%. The model’s mean absolute error (MAE), root mean square error (RMSE), and mean absolute percentage error (MAPE) are 0.658%, 0.780%, and 4.38%, respectively, significantly outperforming conventional machine learning models and traditional assessment methods. This research provides a reliable tool for evaluating pipeline integrity and maintenance planning.

---

Keywords

---