Research On Water Distribution Network Leakage Probability Prediction Method Based On Hydraulic Model

At present,with the accelerated urbanization process,the scale and complexity of China’s urban water distribution network is increasing,and different pipeline materials and diameters coexist in the water distribution network of old and new urban areas,and the leakage problem of water distribution network is serious,and the rapid and effective detection of leakage in large water distribution network has become an urgent problem.With the development of science and technology,a leakage control system based on a hydraulic model to build a leakage detection model to solve the leakage problem of large water distribution network with complex structures and different materials has emerged in the water supply industry,which is of great significance to reduce the leakage rate of water distribution network,improve the efficiency of water supply and guarantee the water safety of residents.In this paper,a hydraulic model-driven online leakage monitoring method for large water distribution network is investigated,and the research is summarized as follows:1.To address the problems of large search space and complex leakage detection in the solution process of large water supply pipeline network,a leakage area identification method based on virtual partitioning is studied.According to the topology of the water distribution network and the location of monitoring points,the network is divided into several virtual areas,and a leaky area identification model is constructed based on the improved Gray Wolf optimization algorithm to identify the leaky virtual areas and estimate the size of the leakage volume in each virtual area.The experimental results show that the model accurately identifies the leakage virtual areas,the average leakage identification error is less than 17.14%,and the change of unit leakage volume is 0.52%of the total water demand,and the detection results provide reference for the field maintenance workers at a later stage.2.To address the problems of complicated process and low accuracy in locating leakage events in the water supply network area,the study predicts the leakage location in the virtual zone,converts the a priori probability of leakage events into a posteriori probability by probabilistic statistical methods,quantifies the uncertainty of modeling error and measurement error in the hydraulic simulation model by using the posteriori probability density function,calculates the probability distribution of leakage pipe sections and leakage volume in the water supply network,and realizes the precise location of The leakage points are precisely located.The experimental results show that the leakage probability prediction model detects leakage at nearby nodes,and can accurately predict the leakage location by collecting multiple sets of data.3.To address the problems of low monitoring and management efficiency and high operation cost of water supply network,we designed a monitoring and alarm system based on Supervisory Control and Data Acquisition(SCADA),which includes a leakage area identification module,a leakage probability prediction module and a network alarm module.By inputting the leakage information from the leakage area identification module to the leakage probability prediction module,and further inputting the results of the leakage location within the predicted area to the pipe network alarm module,the pipe network alarm system is realized to issue alarm information and provide overhaul guidance for water service workers.Figure [40] table [12] reference [49]...