Research On Nodal Demand Calibration And Leakage Detection Of Water Distribution System

Leakage of urban water distribution system can cause waste of water resources,and may also cause water pollution in the network,threatening the safe operation of urban water distribution system.Therefore,timely detection and repair of water distribution system leakage incidents is particularly important.With the continuous development of information technology,how to use intelligent water distribution system management methods to solve the problem of network leakage has attracted much attention.Based on this,this paper conducts the following research around the nodal demand calibration algorithm and leakage detection algorithm of water distribution system:(1)A nodal demand calibration algorithm for water distribution system based on prior information;In order to solve the underdetermined problem caused by nodal demand calibration in water distribution system,current research usually aggregates nodes according to different water use types to reduce the number of unknowns.However,the process of node aggregation is fraught with human subjectivity,and improper grouping can lead to inaccurate calibration results.To solve this problem,this paper uses water consumption information as a priori value to introduce the proposed objective function for calibration,avoiding aggregation of various water consumption nodes.In addition,the impact of monitoring errors and pipe friction coefficient on the proposed calibration algorithm was explored,and calibration suggestions were proposed for the impact of different errors on the algorithm.(2)A wavelet denoising-Kalman filter based leakage detection algorithm for water distribution system.Current data-based leakage detection algorithms are promising,but there is a general problem of low robustness in their use.For example,when traditional Kalman filter is used to detect leakage in water distribution system,the false alarm rate is high due to monitoring errors and random fluctuations in water consumption.To address this problem,this chapter proposes a wavelet denoising-Kalman filter-based leakage detection method for water distribution system.Firstly,a brief introduction is given to Kalman filter-based leakage detection method in water distribution system;then,the basic principles of wavelet denoising and the operational procedures are elaborated;finally,the two algorithms are combined to obtain the final leakage detection algorithm,and the performance of the algorithm is tested in a specific case.(3)A leakage detection algorithm of water distribution system based on nodal demand calibration residual.In recent years,the use of water distribution system monitoring data for leakage detection has been widely concerned by researchers at home and abroad and a variety of leakage detection methods have been proposed.However,there is a common problem in the existing data-driven leakage detection research : these methods can only use a single type of data(such as flow or pressure),cannot combine multiple different types of data for analysis at the same time,and cannot be used in combination with the hydraulic model,which may be the reason for the low robustness of the existing algorithms.In view of the above problems,this section proposes a leakage detection algorithm for water distribution system based on the residual of nodal demand calibration(based on the coupling of data and hydraulic model).The algorithm uses different types of hydraulic information such as hydraulic model,flow and pressure monitoring data to detect leakage,which makes the algorithm more robust.