Extreme Value Distribution Model of Process Pipeline Pitting Depth

A method for constructing an extreme value distribution model of pitting depth in process pipelines was introduced. Using the measurement results of pipe wall thickness reduction (pitting depth) as the dataset, the generalized extreme value (GEV) distribution was used to fit the thickness reduction dataset. Based on the parameter estimation results, a pitting depth extreme value distribution model was established. Finally, the model was validated. The results indicate that the root mean square error of the cumulative probability of the model was 9.5%, indicating a high accuracy of the model.