Study On Layout Optimization Of Monitoring Points For Pipe Burst Detection In A Water Supply Network

The pipe burst of the water supply network(WSN)is an emergency.Once a pipe burst event occurs,it usually has a serious adverse impact on society,which is one of the safety issues that water supply management departments focus on.The installation of monitoring equipment,such as pressure gauges,is a significant means of burst event detection.The construction and maintenance costs of a monitoring network constrain the number of monitoring points.Optimising the spatial location and number of monitoring points can improve the capability of the monitoring network to detect burst events.Firstly,based on the pressure and flow data collected by the SCADA system,the natural fluctuation characteristics of the WSN are analyzed.The research found that not all bursts can be detected.A pipe burst event can be detected only when the effects of the burst are distinguishable from background noise in the monitoring network.Secondly,the pipe network hydraulic model is used for burst simulation analysis.In order to evaluate the burst event,the concept of burst intensity is proposed.After analysis,it is recommended to use the pipe velocity increase to represent burst intensity In order to facilitate the burst simulation,the method of specifying the burst intensity is adopted.Through the analysis of three cases,it is found that the results obtained by using different hydraulic models for burst simulation are the same,but the pipe velocity increase and the pressure background noise threshold have influence on the simulation results.Thirdly,in order to prioritize leak detection in the most important pipes and maximize the monitoring network capability to detect water leakage in a WSN,the spatial optimization function defined under the specified burst intensity is established.In order to evaluate the detection capability of monitoring points on the WSN,the evaluation index of pipe detection coverage is established.The marginal effect in economics is used to determine the number of monitoring points,and the genetic algorithm is used to optimise the spatial location.The layout optimization method of monitoring points for pipe burst detection in a water supply system is put forward systematically.Finally,a comparison of the method we developed with other methods of burst event detection described in the literature shows that the method is superior to other methods.The optimization method can be well applied to cases of different scales.The calculation results show that the monitoring points are evenly distributed,which are consistent with intuition.The pipelines that can be detected by the monitoring network are clear,which is helpful for the water company to know the detection capability.