Citation

Zheng D, Cheng Y, Li H, He T. China Saf. Sci. J. 2022; 32(1): 110-117.

Copyright

(Copyright © 2022, China Occupational Safety and Health Association, Publisher Gai Xue bao)

DOI

10.16265/j.cnki.issn1003-3033.2022.01.015

PMID

unavailable

Abstract

In order to achieve more accurate prediction of CO2 corrosion rate of oil gathering and transportation pipelines, considering the low prediction accuracy of original GRNN, an improved swarm intelligence algorithm was proposed to optimize it. Then, GRNN model, AFSA optimized GRNN (AFSA-GRNN) model and adaptive improved AFSA-GRNN (IAFSA-GRNN) model were used to predict CO2 corrosion rate of X65 pipeline steel. The results show that the prediction accuracy of GRNN model is greatly improved by using AFSA and IAFSA to optimize smoothing factor S, with its mean relative error of prediction results being reduced from 36.09% to 7.20% and 6.90%, respectively. Compared with AFSA, IAFSA optimized GRNN has not only higher prediction accuracy, but also faster convergence rate. The AFSA-GRNN converges at the 164th iteration, while the IAFSA-GRNN does that at the 109th one, indicating that AFSA could improve convergence rate of optimization process and further improve prediction accuracy of GRNN through adaptive optimization. © 2022, Editorial Department of China Safety Science Journal. All rights reserved.

Language: zh