Research On Optimal Sensor Placement For Water Loss Monitoring Based On Uncertainty Analysis In Urban Water Distribution Networks

With the continuous increase of urban population,the problem of water shortage has become increasingly prominent.What's more,the current water supply network leakage rate is still high.Thus,it is imperative for water companies to establish a perfect urban water supply network monitoring system,aming to monitor of leakage events continuously and efficiently.As a consequence,reducing the leakage rate and ensureing the water stable of end-users.Considering the high cost of construction,it is especially important to configure a sensor network within the project cost budget that can improve the fault diagnosis and leakage locating accuracy.Urban water distribution networks'efficient and stable water supply will eventually be realized.For the foregoing requirements,this dissertation focuses on the sensor optimal placement problem of urban water distribution network.The main work and innovations are as follows:1.This paper apply the Latin Hypercube Sampling(LHS)method to analysis the uncertainty of the hydraulics model of water distribution networks so that the diagnosis effect of existing leak detection method won't easily affected.In this part,we analysis the sources of the various uncertainties affecting the model accuracy and gives the treatment methods.The impact of the uncertainty of model parameters on the output value is quantified.The nodes that are easily affected by the uncertainty of the operating condition of the network are identified,and the layout of the monitoring points is instructed.In this study,the real pipe network system example is used to discuss the influence of water fluctuation on the system pressure distribution at each node,which verifies the feasibility and practicability of the proposed method.2.In this paper,an optimal sensor placement for water loss monitoring based on uncertainty analysis in urban water distribution networks is proposed.This strategy can markedly enhance the robustness of the monitoring system.LHS based uncertainty simulation of actual operation of the network is researched in this part.Subsequently,this paper suggests a two-objective optimization function by which can characterize the identifiability of leakage,and then turn the optimal sensor placement problem(OSP)into a multi-objective problem.For the model of sensor placement optimization,a solving method based on non-dominated sorting genetic algorithm ?(NSGA-?)is designed.Due to the imprecise nature of the decision maker's judgment,the fuzzy set theory is applied to derive efficiently the optimal solution for decision makers.The research verifies the feasibility and effectiveness of the proposed multi-objective optimization strategy.under actual operation conditions.on different time scales.Except that,the performance changes of the algorithm under different time scales is also studied.3.In this part,a set of sensor placement strategy for new nodes under sensors already installed in the network is proposed.Then,provide scientific and rational expansion plans in water supply pipe network monitoring system for water companies.And on this basis we also develop an investment-precision relationship indicator to evaluate the sensor placement plans.On the basis of previous analysis,the layout optimization of new nodes under different model uncertainties is carried out.Note that the assessment is build on a real network situated in a southern Chinese city,meaning that the analysis of the science and feasibility of former method is trustworthy.After studying the tendency of monitoring system performance under different model uncertainties,the robustness of the layout strategy and the cost-benefit function under complex conditions are verified.In general,the proposed strategy could reduce the effect of uncertainty in water distribution networks to leakage detection effectively.And on the basis of network layout strategy and cost-benefit evaluation system,it can not only lay methodological basis for efficient operation of the leakage monitoring system,but also provide decision support for practice of engineering.