Abstract: In order to study the influence of the initial condition for unequal interval grey models(UGM(1,1)) on pipeline corrosion prediction, firstly, the modeling steps of unequal interval pipeline corrosion prediction grey model under different initial conditions were discussed, secondly, aiming at solving the problem of selection of initial conditions in the traditional unequal interval pipeline corrosion prediction grey model, a sliding unequal interval grey pipeline corrosion prediction model (SUGM(1,1,ρ)) under a sliding initial condition was built up by combing the ideas of metabolism and new information priority principle. Finally, the models were used to predict the pipeline corrosion of a certain marine riser as an example, and a comparison was made between the prediction effects of SUGM(1,1,ρ) model and the unequal interval pipeline corrosion prediction grey models under different initial conditions. The effectiveness of SUGM(1,1,ρ) model was verified. The results show that the average relative errors of predicted results obtained by using four unequal interval pipeline corrosion rate prediction grey models under different initial conditions are 3.16%,3.35%,3.49% and 3.36%, and that obtained by using SUGM(1,1,ρ) model is 1.77%.

Key words: unequal interval grey model(UGM(1,1)), initial conditions, metabolism, sliding modeling, corrosion prediction