Design And Research Of Urban Water Supply Network Monitoring And Management System

With the continuous urbanization and rapid development of social economy in China,the scale of urban water supply network is expanding and the structure of old and new pipe networks is complex.With the continuous promotion of intelligent management,new challenges have been put forward to the research of urban water supply network monitoring and management system.In this dissertation,the research is carried out in the following aspects: analysis of abnormal data of water supply network pressure and flow and early warning strategy of pipe burst;database optimization and data visualization;design and development of urban water supply network monitoring management system.(1)In view of the problems of false alarms and omissions in the monitoring management system of urban water supply network when burst pipes and leaks occur in water supply network,this dissertation proposes an abnormal data classification model based on BP neural network to realize the classification of normal data,burst pipe data,under-pressure data and over-pressure data of water supply network.The model is compared with Multi-layer Perceptron(MLP)algorithm,K-Nearest Neighbor(KNN)algorithm,Random Forest(RF)algorithm,Support Vector Machine(SVM)algorithm based on the same data set.The experimental results show that the model can achieve higher classification accuracy,provide technical support for the identification of abnormal data of urban water supply network,and provide graded warning according to the classification results,and reduce the false alarm and leakage rate compared with the traditional alarm methods.(2)With the continuous accumulation of pressure and flow history data in the database,the performance of the data query part is reduced.In order to improve the system performance and increase the response time of the system,the database optimization technology is studied,mainly taking the methods of data partition query and index optimization,and the experiment proves that these two methods can effectively improve the query performance.In order to improve the visualization of the system,this dissertation applies the ECharts tool to show the data transmitted by the data collection terminal in real time and the historical data in the form of line graphs,so that the managers can see the data trends of the pipe network and the equipment information more intuitively.(3)Finally,a monitoring management system for urban water supply network is developed,including functional modules such as login function,data query and equipment query function,map management function,early warning function,system security and management function and security function.Spring Boot +My Batis is used as the system development framework,Java is used as the development language,and the neural network model is built based on Py Torch framework;My SQL database is used for data storage;Netty framework is used as the communication framework between the data collection terminal and the platform;ECharts tool is applied to realize the data visualization function Netty framework is used as the communication framework between the data collection terminal and the platform;ECharts tool is applied to visualize the data;and database optimization is achieved through database splitting and index optimization to improve the efficiency of data query.