Study On Leakage Law Of Pressure Water Supply Pipeline Based On Different Leak Failure Forms

The study found that physical leakage is an important part of the loss of the pipeline network.It not only causes economic losses to the water supply company,but also wastes mounting precious fresh water resources.How to accurately predict physical leakage in the urban pipe network is an important issue.However,there are still some shortcomings in the present existing various leakage models and researches.Therefore,starting from the point-leak theory,this thesis conducts leakage test experiments on different leakage forms of pipelines,and then analyzes the impact of leakage forms to the pipeline pressure fluctuations and leakage volume.The traditional point-leak index model and the derived leakage formula only consider the correspondence between the leak loss and the leak area,the pipeline pressure.There is no quantitative analysis of the relationship between the leak loss and the leak shape.In this paper,the three types of leakage,namely the lengthwise,the circumference and radial,are studied.The software FLUENT on computer is used to simulate the leakage.The relationship between the leak shape and its pressure field,and velocity field,and the leakage volume are analyzed theoretically.Determine its necessary to research.Then through a combination of physical experiments and computer numerical simulation,a large number of leak experiments were carried out to investigate the effects of the leak shape,the leak area ratio and the leak-water pressure on leakage loss volume when the leakage reached to a steady state.The leakage variation coefficient k in the traditional leakage index model is refined,and the specific expression formula between the leakage coefficient k and the leak area ratio s of the exponential model under different failure forms is obtained,and the traditional leakage index model is therefore optimized.Then the new model formula expression has a highly goodness of fit.The establishment of the new leak index model not only can more accurately reflect the leakage situation in the pipeline,but also provide a basis for the calculation of pipeline leakage loss volume under different leak forms.Based on the leak experimental data completed in the physical experiment,the effects of different leak shapes on the transient pressure fluctuation volatility and leakage loss of various pipes,different pipe diameters and different pipeline pressures are summarized.BP neural network is established.The network,through a large number of input and output training,debugs the optimal structural composition of the neural network,and establishes a nonlinear mapping relationship between the pressure fluctuation data at the leak and the leak forms and the leakage volume of the pipeline.Through testing and verification,the BP neural network can make accurate predictions for the leak experiments.This experiment not only provides a feasible method for predicting the leakage loss in the water supply network in the future,but also establishes a connection between the forms of the leakage point and the pressure fluctuation around the leak point,and diagnoses the leak forms from the pressure fluctuation data in the future.From the aspect of the leak forms,it could possibly analyzes the cause of leakage and provides a new way to effectively control the leakage loss of the pipe network from the roots.