Study On Layout Optimization Of Monitoring Point In Water Distribution System Based On Evolutionary Algorithm

The pipe burst not only causes serious waste of water resources,but also causes pollution of water supply system,so it is of great significance to find and repair the burst pipe in time.The water pressure monitoring points in the water supply network can obtain the node pressure situation,and it is an important means to detect the burst pipe according to the abnormal pressure signal.Due to the limitation of funds and other conditions,t he number of monitoring points is limited,so it is necessary to optimize the layout of monitoring points.In view of this,this paper develops the study on layout optimization of monitoring points in water distribution system based on evolutionary algorithm.The results show that:(1)Study on hydraulic simulation analysis of burst pipe.Firstly,combined with the principle of pipe network hydraulic calculation,the variation characteristics of section head before and after pipe burst are analyzed.Secondly,the pipe burst simulation method in the current research is discussed,and the pipe burst simulation method of this study is determined.Then,taking three pipe networks of different scales as research es,the impact of pipe burst is simulated under different burst flow and background noise threshold.The results show that the impact range of the burst pipe on the pipe network is affected by the burst pipe flow and the background noise threshold.(2)Study on monitoring point optimization model construction for pipe burst detection and computation.The mathematical model of monitoring point optimization is established by minimizing the number of monitoring points and maximizing the leakage of burst detection,and a multi-objective non-dominated differential evolution algorithm is improved.In order to verify the superiority of this algorithm,the methods of solving monitoring point optimization model based on single objective genetic algorithm,multi-objective non-dominated sorting genetic algorithm and multi-objective non-dominated differential evolution algorithm is constructed respectively.Taking three pipe networks of different scales as research es,the results show that the solution solved by the nondominated differential evolution algorithm has be tter quality and higher computational efficiency,and is more suitable for solving the problem of monitoring layout optimization in water supply pipe networks.(3)Study on the improvement of evolutionary algorithm performance improved by clustering preselection method.Firstly,the system clustering algorithm is used to classify the nodes basing on the node pressure change matrix.Secondly,according to the Euclidean distance,the center nodes of various types are pre-selected as monitoring nodes.Then,the layout of monitoring points is further optimized by evolutionary algorithm.Taking three pipe networks of different scales as researches,the results show that the clustering preselection method can make the algorithm find the layout schemes with larger burst pipe coverage,which significantly improves the performance of the evolutionary algorithm in solving the problem of monitoring layout optimization in water supply pipe networks.(4)Study on the optimal number of monitoring point based on cost-benefit analysis and the analysis of burst pipe detection characteristics.Firstly,the investment cost and burst coverage of different monitoring points are normalized by the special function.Secondly,the net cost is calculated according to the investment co st of the monitoring point and the uncovered rate of the burst pipe.Then,the optimal number of layout is determined by comparing the net cost of different monitoring points.Finally,the characteristics of pipe burst detection are analyzed for the optimal number of monitoring layout scheme.Taking three pipe networks of different scales as researche,the results show that the cost-benefit analysis method can effectively determine the optimal number of monitoring point,and the distribution of monitoring point is relatively uniform,which achieves a better effect of burst pipe detection.