Study On Optimal Operation Of Secondary Water Supply Based On Hydraulic Model And Genetic Algorithm

In order to orderly promote China’s carbon peak and carbon neutralization,"energy conservation and carbon reduction"has become the key word of construction and development in various fields.Secondary water supply is a pressurized water supply method to compensate for the insufficient water supply pressure of the municipal pipe network and ensure the water consumption of users in high-rise buildings.Variable frequency drive has been widely used in secondary water supply,but its energy-saving role has not been fully played.In order to further tap the energy saving potential of secondary water supply system and improve the energy efficiency level of urban water supply,this paper studies the optimal scheduling of secondary water supply system in residential areas of Hefei,and the main work are as follows:Water volume prediction is the premise of optimal dispatching,and the water volume law analysis and prediction of DMA can guide the estimation of water supply pressure in the pump house.Based on the historical data of the water meter,this paper analyzes the change rule of the water supply in the residential quarter:the daily water supply in summer is significantly higher than that in winter,and the 24-hour water supply changes in a"M"shape,there are the early water consumption peak period and the late water consumption peak period,which are 6:00-8:00 and 20:00-22:00 respectively in June.Three short-term water supply time series prediction models are constructed,including the LSTM（long short-term memory）model,the SVR（support vector regression）model,and the Transformer model,which is applied to water supply prediction innovatively.Predicting the water supply volume in different DMAs for three consecutive days,the results are as follows:The best prediction effect is using the Transformer model in residential area A,where with the smallest occupancy rate,MAPE=10.537;The best prediction effect is achieved by using Transformer model in residential area B,MAPE=10.197;and in residential area C,where with the highest occupancy rate,the best effect is to use LSTM model,MAPE=6.821.Hydraulic modeling is an important step for optimal operation.This paper collects the basic data and operation monitoring data of the water supply network and secondary water supply pump house in the research area,establishes the Off-line microscopic hydraulic model of water supply network through the software EPANET.Then verifies the model,so that the average error between the simulated pressure and the actual pressure at the pressure monitoring point is controlled below 15 Kpa,which meets the recommended value of the industry’s accuracy standard.Based on the water volume research and hydraulic model,this paper studies the optimization of variable pressure water supply in residential pump houses and the joint optimal operation of regional pump tanks.The variable-pressure water supply scheme based on water use law and the variable-pressure water supply scheme based on end pressure feedback are proposed.The hydraulic simulation results show that the two schemes save about 4.6%and 15.25%of electricity respectively.The variable pressure water supply test run is adopted in some residential areas in Hefei.Compared with the historical constant pressure operation results,the Yungu residential area with high occupancy rate has a better energy-saving effect:the daily average power consumption decreases by 9.85%,about 35.2 k W·h,and the annual production cost of the pump house can be saved by about 74100 yuan based on the electricity price of 0.5853 yuan/（k W·h）.The feasibility and energy saving of the scheme are verified.Establish an joint optimal dispatching model of secondary water supply pump and water tank,transform the constrained problem into an unconstrained problem through penalty function,and then use genetic algorithm to solve the model.Compared with the actual operation plan,the optimized dispatching plan generated by the solution increases the node pressure of the regional pipe network by 0.5～2 m H₂O during the late peak period of water supply（after20:00）,and the daily power consumption of the secondary water supply pump house in the region decreases by 9.53 k W·h,which alleviates the contradiction between supply and demand while saving energy and reducing consumption.Figure[48]Table[27]Reference[71]...