Risk assessment method of urban gas pipeline based on GA-PSO-XGBoost model

Objective Aiming at the limitations of the traditional risk assessment methods in the risk assessment of urban gas pipelines, such as the difficulty of quantitative analysis and strong subjectivity, a gas pipeline risk assessment model based on machine learning is proposed.

Method Based on the genetic algorithm (GA), particle swarm optimization (PSO), and eXtreme gradient boosting (XGBoost) algorithm, the GA-PSO-XGBoost urban gas pipeline risk assessment model is constructed. The model uses the GA-PSO hybrid algorithm to optimize the hyperparameters of XGBoost, thereby improving the accuracy of urban gas pipeline risk assessment.

Result The model test results show that the mean square error (MSE), mean absolute error (MAE) and coefficient of determination (R²) of the GA-PSO-XGBoost model are 4.1264, 1.4939 and 99.1146%, respectively, which verifies the high accuracy and reliability of the model.

Conclusion The GA-PSO-XGBoost model provides a new method for the risk assessment of urban gas pipeline, which has an important application value for improving the safety management of gas pipeline.