The Research And Application Water Distribution Pipeline Leakage Prediction

Leakage of water distribution systems is a serious problem in water industries. Leakage not only wastes precious water resources, but also leads to much financial loss. Leakage of water distribution systems has become an important issue of concern to study in the world.Based on collecting leakage data and making leakage investigation, according to summarize the reason for the network leakage, management measures were proposed, and also the technical measures for reducing leakage frequency and leakage quantity were presented from the view point of design, construction and maintenance.According to characteristics of leakage of water distribution systems, the gray GM(1,1) prediction models, second order multinomial exponential smoothing models and auto-regressive moving average (ARMA) models have been chose to establish, and are applied in the leakage quantity forecast. Research results show that application of these methods is feasible and effective, but they are relatively difficult to guarantee the precision. As a result of that, the paper comes up with combination forecasting models and elaborates basic idea of combination forecasting models, and make variance- covariance optimal combined models which combined of three single modles applied in the leakage frequency forecast of water distribution pipeline. Results show that the combined models can synthetically use the information that is provided by every single models, the shortage of every single model can be avoided, the combined model have high precision.Combination forecasting models for leakage of water distribution systems not only help to predict the trends of leakage in the future, but also provide the basis of analysis on leakage control and maintenance of water distribution systems. It is in favor of maintenance and reconstruction of water distribution systems, has instructive significance for leakage control and saving water resource.