Abstract: A new grey prediction model based on PSO optimization algorithm was proposed for the prediction of seabed pipeline corrosion. Based on the traditional gray GM (1,1) model, the PSO algorithm was introduced to optimize the background weight λ and the model was dynamically updated by introducing equal dimension grey recurrence method. RGM (1,1) and RPGM (1,1) were applied to the prediction of submarine pipeline corrosion. Comparing the prediction results of the three models, it was found that the grey prediction theory was suitable for the prediction of submarine pipeline corrosion,and the prediction result of the RGM(1,1) model was slightly better than that of the traditional GM(1,1) model. The pre-measurement accuracy of the RPGM (1,1) model was greatly improved compared to the other two models. The PSO algorithm had a significant improvement effect on the traditional model, and the RPGM (1,1) model has high engineering application value.