Font Size: a A A

Research On Operation Of Water Distribution Network Aiming To Energy Saving

Posted on:2015-05-31Degree:MasterType:Thesis
Country:ChinaCandidate:G R ZhaiFull Text:PDF
GTID:2272330422491611Subject:Municipal engineering
Abstract/Summary:PDF Full Text Request
In the water distribution system power consumption accounts for a large proportionin the cost of operation.In2005,annual power consumption of the national waterindustry reached to40billions kwh. Water supply scheduling in china is still in the stageof experience,which has great blindness and don’t have the property of energy saving.Water supply scheduling has great energy-saving space, it is necessary tostudy the watersupply scheduling.This paper studys water supply scheduling mainly from three aspects:forecastwater consumption、establish hydraulic model、establish and solve scheduling model.Firstly,this paper studies the forecast of24hour consumption.In the begin,thispaper uses the pure time series ARIMA model to predict the water consumption.Theprediction accuracy is about5%,which is not very good.Then this paper divides thewater consumption of24hours into the water consumption of a day and the percentageof a day and predict separately.Using the SVM method to predict the day waterconsumption obtain good results,the prediction accuracy is about5%.Forecasting thethe percentage via PSO-BP,which achieves a good result,the prediction accuracy isabout3%.Finally,the prediction accuracy of the hour water consumption is in5%.Secondly,the paper introduces the establishment of the microscopic model andmacroscopic model.The microscopic model established doesn’t meet the requirement ofthe water supply scheduling. Later this paper attempts to establish the macro hydraulicmodel,which shows the mapping relationship of factory pressure,factory flow,monitoring vaule of each pressure monitoring point.This paper establishes thesupply-demand macro hydraulic model and demand-supply macro hydraulic modelseparately using PSO-BP,which achieves the accuracy of5%for the data currently onhand.Then,this puts forward the pure numerical of pump checking,and solves theproiblem via Cuckoo search algorithm.it obtains good results.The error of flow isbetween100and200m3/h.Finally,the paper establishs the scheduling model based on macroscopic model,andsolves the model with PSO-BP aiming to obtain the lowest residual pressure of pressuremonitoring,getting the operation pressure of24hour.In general the solution is1m lowerthan the actual operation pressure.It has great energy saving space,but needs to bevalidated.At last,this paper re-designs the pump scheme,which is energy saving and alsoimprove the security of water supply.
Keywords/Search Tags:Water consumption forecast, Arima model, PSO-BP Network, SVM, Particle swarm optimization algorithm, Macro hydraulic model, Pumpcheck
PDF Full Text Request
Related items