Inverse transient analysis for burst location in water pipelines based on particle swarm optimization

Currently, existing methods for burst location in water transmission pipelines generally suffer from insufficient positioning accuracy, making it difficult to quickly and accurately identify the burst location, size, and occurrence time. Therefore, this study establishes a burst transient flow model that couples the Method of Characteristics (MOC) with the Trikha-Vardy-Brown (TVB) unsteady friction and conducts a characteristic analysis of burst parameters. A burst experiment platform was built to verify the model’s accuracy. The burst location problem is transformed into an optimization problem, and an objective function including the burst location, size, and occurrence time is constructed. Four optimization algorithms were used to solve the objective function. In addition, by coupling the burst location and occurrence time to achieve dimensionality reduction, an improved PSO (Particle Swarm Optimization)-ITA (Inverse Transient Analysis) pipe-burst localization model is proposed, and parameter sensitivity analysis is carried out. Its robustness was verified in a simulated water pipeline system with branch pipes. The results show that the coefficient of determination (R²) of the burst transient flow model under different tests was close to 1, indicating the model's accuracy. Changes in the burst parameters (burst location, size, and occurrence time) have obvious effects on pressure fluctuations and fitness functions. The PSO algorithm had the highest accuracy in solving the objective function. The calculation accuracy and efficiency of the improved PSO-ITA pipe-burst localization model after dimensionality reduction were significantly improved, and it was unaffected by the pipe-burst location, position, and number of measurement sections. The model can accurately calculate the burst location and occurrence time when the pipeline friction coefficient is inaccurate. When the pipeline wave velocity varies within a certain range, it can accurately calculate the burst location, size, and occurrence time. The improved PSO-ITA pipe-burst localization model exhibited good robustness in complex water pipeline systems with branch pipes.