Research On Campus Water Resources Usage Forecast And Optimal Dispatch

With the rapid development of the economic level and the improvement of people's living standards,the consumption of resources is gradually increasing,and the total amount of resources is limited,which also causes various resource allocation problems.Among them,the contradiction of water resources is particularly prominent in these problems.The uneven distribution of water resources and the different demands for water resources in various regions have caused the contradictions between water supply and demand in various regions to become increasingly prominent.Therefore,the rational allocation and use of water resources is particularly important.In order to alleviate these contradictions and meet people's demand for water,a rational planning and design of water supply is needed.In the water supply network system,the amount of water supply should be based on the water demand of people.Therefore,accurate and efficient prediction of people's water consumption should be obtained and the use of water resources should be reasonably dispatched.Correct prediction and efficient allocation are helpful for water supply,water use,and water conservation,as well as for rational planning and the effective use of water network systems.This paper proposes a method for forecasting water demand in the interval of sparse deep confidence network combined with wavelet decomposition,and predicts water demand in the future based on historical water consumption data.Most water demand forecasting methods generally directly predict water demand,but in fact,water demand also has certain regularity.Discovering and digging out these laws should be of great help to its prediction.Therefore,this paper uses wavelet transform to decompose the water consumption data to obtain its hidden periodic,trend and random terms,which makes the prediction process more intuitive,the prediction training of the model is more efficient,and the daily water fluctuation range on campus There are also very accurate predictions,the accuracy of which can reach more than 95%,which can provide a reliable basis for the subsequent optimal scheduling of water resources.Specifically,the method of water demand forecasting is firstly introduced,the classic water demand forecasting methods and related problems are introduced,and then to overcome these shortcomings,a sparse deep confidence network combined with wavelet decomposition is proposed to carry out water demand decomposition,using confidence level for interval prediction,and also compared with the classic water demand prediction methods in the past.Analysis of the final forecast results shows that the water demand forecast value curve basically fits the actual water consumption data,and the average relative error is also controlled within 2%,and the maximum error is only 4.2%.As can be seen,the method we proposed is more accurate in the prediction of water consumption,and there will be no large deviation.Then,based on accurate water demand forecasting,water use scheduling is performed.First,a mathematical model of water resources scheduling is established,and comprehensive functions such as population goals and water environment goals are determined as the objective functions.Water supply,water demand,and water use are the objective functions.Constraints such as cost are used as model constraints.The proposed dual-parallel genetic algorithm is used to optimize the parameters of the water supply pump to achieve optimal water use scheduling.The water resource allocation method optimized by this algorithm can not only meet the daily water demand of teachers and students,but also realize the reasonable allocation of limited water resources,reduce the water supply pressure,and meet the water demand of more people.Compared with the traditional method of water use,the method of water resource allocation obtained through optimization no longer blindly meets the single demand for water,but instead it is considered more extensively,which can meet the current needs and generate more value from limited water resources.It can also prevent people from paying less attention to water resources and wasting water due to the sufficient water supply.In addition,by installing a watersaving device in the pipeline,the optimal use of water resources is further realized after the use of a dual-parallel genetic optimization algorithm to optimize the allocation of water.Finally,the actual water consumption data of the campus shows that this method of sparse deep confidence network combined with wavelet decomposition and campus water demand forecasting and dual-parallel genetic optimization algorithm is accurate and reliable for water resource allocation,which provides favorable technical support for the effective water resource utilization.