Corrosion depth prediction of pipeline based on improved gray model

Abstract

Establishing an accurate prediction model of pipeline corrosion depth is of great practical significance for effectively evaluating the service life of pipeline, which has the advantages of saving manpower and financial resources. Based on the basic theory of gray GM (1,1) model, the improved GM (1,1) models were established by using the method of background value optimization (improved model I), translation transformation (improved model II) and combination of background value optimization and translation transformation (improved model III), and the differences of prediction accuracy between the improved models and the traditional model were analyzed. The results show that the mean relative errors of improved models are lower than those of traditional model. When the coefficient C is 30, the mean relative errors of improved model II and improved model III are only 1.073% and 1.097%. The accuracy of each improved model is higher than that of the traditional model, and the accuracy of improved model III is generally high. With the increase of coefficient C, the prediction accuracy of improved model III and improved model II gradually increases, the reasonable selection of coefficient C is helpful to improve the accuracy of improved model II and improved model III.