基于FCM聚类和漏失模拟的给水管网压力监测点布设  

Optimized layout of pressure monitoring points in water supply network based on FCM clustering and leakage events simulation

作　　者：徐志强 刘崎峰 刘佳峰 单黎明 裴文思 XU Zhiqiang;LIU Qifeng;LIU Jiafeng;SHAN Liming;PEI Wensi(College of Ecology and Environment, Inner Mongolia University, Hohhot, Inner Mongolia 010021, China)

机构地区：[1]内蒙古大学生态与环境学院,内蒙古呼和浩特010021

出　　处：《排灌机械工程学报》2022年第6期642-648,共7页Journal of Drainage and Irrigation Machinery Engineering

基　　金：内蒙古自治区成果转化引导项目(CGZH2018095)。

摘　　要：为了提高对供水管网压力监测的可靠性,以北方某城镇供水独立计量分区为模拟研究对象,提出一种结合模糊C均值(Fuzzy C-Means,FCM)聚类和对模拟漏失事件感知数量最大化的给水管网压力监测点优化布设方法.首先,基于水量变化的节点压力敏感系数矩阵,利用FCM聚类算法对管网节点进行聚类,在各子类中预选出备选压力监测点;然后,随机模拟可能发生的漏失事件测试符合条件的不同优化组合的备选监测点对漏失事件引起压力变化的感知情况,确定优化的监测点布设方案.案例研究结果表明,该方法确定的压力监测点布设方案能够在保证至少2个监测点感知到同一个漏失事件的前提下,以最少3个监测点感知到96.36%以上的漏失事件.该方法在提高监测可靠性的同时可以节约监测点布设成本.In order to improve the reliability of pressure monitoring points in the water supply network,a layout method of pressure monitoring points combining fuzzy C-means(FCM)clustering with the maximum perceived number of leakage events was proposed based on a District Metered Area(DMA)in a town in northern China.Firstly,the FCM algorithm was used to cluster all nodes based on the nodes pressure sensitivity coefficient matrix of water quantity change in the pipe network,and the alternative pressure monitoring points in each subclass were preselected.Then,the possible leakage events were simulated randomly to test the perception of the alternative monitoring points with different optimized combinations for the pressure changes caused by the leakage events,and the optimal monitoring points′layout scheme was determined.The simulation results of this case show that more than 96.36%of the leakages were detected with at least three monitoring points on the premise ensuring that at least two monitoring points can perceive the same leakage event.This method can improve the monitoring reliability and at the same time save the cost.

关 键 词：压力监测点 优化布设 模糊C均值聚类 漏失模拟 给水管网 

分 类 号：TU991[建筑科学—市政工程]