摘要
目前二次供水系统中水泵普遍存在“大马拉小车”导致能耗偏高的问题,为了解决这一问题,需要建立一种更加贴合实际工况的二次供水流量预测方法。文章以江苏常州地区149个二次供水小区的流量监测数据为基础,形成二次供水流量特征样本集,首次综合增压户数、入住率、最高日流量、最高日最大时流量等特征数据,依靠基于遗传算法优化的BP神经网络进行数据挖掘,构建具备可用性的二次供水流量预测模型,并应用于老旧泵房的水泵改造工作,电耗降幅达19%以上,取得了良好的节能效果。流量预测模型可以为二次供水设计选型、节能改造等工作提供更精准的流量评估工具,也为二次供水节能减排提供新的研究思路,助力实现“双碳”目标,推进供水绿色发展。
At present,the water pump in the secondary water supply system generally has the problem of high energy consumption caused by“big horse pulls a small carriage”.In order to solve the problem,a kind of secondary water supply flow prediction model which is more suitable for the actual working conditions need to be established.Based on the flow monitoring data of 149 secondary water supply communities in Changzhou,Jiangsu Province,this paper formed a characteristic sample set of secondary water supply flow.It was the first time that the characteristic data such as the number of pressurized households,the occupancy rate,the highest daily flow rate and the maximum hourly flow rate on the highest flow rate day were integrated,and the BP neural network optimized based on genetic algorithm was used for data mining.The prediction model of secondary water supply flow with availability was constructed and applied to the renovation of old pump house.Then power consumption had decreased by over 19%.Good energy saving effect was achieved.Flow prediction model can provide more accurate flow evaluation tools for secondary water supply design selection,energy saving renovation and other work,but also provide new research ideas for secondary water supply energy saving and emission reduction,help achieve the“double carbon”goal,and promote the green development of water supply.
作者
肖磊
蒋瑜
刘书明
吴雪
陈春芳
XIAO Lei;JIANG Yu;LIU Shuming;WU Xue;CHEN Chunfang(School of Environment,Tsinghua University,Beijing 100084,China;Changzhou CGE Water Co.,Ltd.,Changzhou 213003,China)
出处
《净水技术》
CAS
2024年第5期71-79,共9页
Water Purification Technology
基金
国家水体污染控制与治理科技重大专项(2017ZX07201002)。
关键词
二次供水
遗传算法
BP神经网络
流量预测
节能减排
secondary water supply
genetic algorithm
BP neural network
flow prediction
energy saving and emission reduction
