Research And Application Of Intelligent Scheduling Algorithm For Multi-type Water Sources Of Urban Water Supply

Smart city construction has become an important aspect of China’s practice of supply side reform and an important standard to measure the sustainable and healthy development of cities and towns.As an important part of smart city,smart water construction relies on new technologies such as Internet of things,big data and cloud computing.It can not only establish a scientific urban water management system,but also effectively improve water efficiency and save a lot of water resources.The optimal operation of water supply system is an important link in the construction of urban intelligent water affairs,and the accurate and efficient prediction of urban daily water consumption is the prerequisite for the optimal operation,which can provide a scientific basis for the follow-up operation decision-making of water supply units.Taking the actual water use system in an area of Xi’an,Shaanxi Province as an example,this paper studies the short-term daily water consumption prediction problem and multi type water source scheduling problem,and develops a practical information system for operation auxiliary decision-making.For the short-term daily water consumption prediction,firstly,it is found that the prediction accuracy of the traditional machine learning prediction method is not high in practice through experiments.Then,according to the historical water supply data and actual operation status of the local water company,two influencing factors with higher correlation with daily water consumption and including the time series characteristics of daily water consumption data are introduced as improvement schemes.After applying it to three common machine learning algorithms,the performance is compared from all aspects.The final experimental results show that the scheme can significantly improve the accuracy of water volume prediction,and the time complexity of the algorithm does not increase significantly,which can meet the requirements of enterprise application.On the premise of accurate water demand prediction,this paper explicitly considers the urban water supply operation rules such as water distribution modes,the start and stop of well group pump units and water quality requirements,establishes a multi-objective optimization model for raw water distribution and well group start and stop scheduling including multiple types of water sources,and uses the elitist nondominated sorting genetic algorithm(NSGA-Ⅱ)for optimization.On the premise of ensuring the quality of water supply,compared with the traditional manual scheduling,the optimized scheduling strategy has lower resource consumption and water supply cost,and the water supply structure of the whole city is more reasonable.Based on the above research and overall requirements of the project,a corresponding urban intelligent water management system is developed.This system mainly includes user management,authority management,site management,supervisory control and data acquisition(SCADA),district metered area(DMA),water demand prediction and other modules.The database is designed according to the correlation between various data.The back end adopts spring framework and Java language to realize logic and algorithm,and the front end adopts Vue+elementUI for code development to realize the system development of B/S architecture.The urban intelligent water management system can realize the real-time sharing of water information,which makes the decision-making of water supply units more convenient and efficient,and has certain practical significance.