Study On Campus Water Demand Forecasting And Scheduling Based On Neural Network

The increase of population leads to the increase of water consumption,and some areas still have the situation that supply and demand are difficult to balance,which leads to a significant decline in the application rate of water consumption.Based on the above,reasonable application of water resources is particularly important,and reasonable water demand forecasting and water scheduling is an effective means of rational application of water resources.Campus water demand forecasting and water scheduleing have been hot topics in recent years.Due to the densely populated campus and large water demand,the campus' s water demand forecasting and water supply schduling can be used to timely discover anomalies in campus water use and issue them reasonable water use indicators,rational formulation of water resources policies,rational formulation of water supply policies,application of its methods and ideas to other areas,the purpose of avoiding waste of water resources,and promoting the healthy development of society and the economy.In view of the uncertainty caused by many factors in campus water use,this paper proposes a campus water demand forecasting model based on the gray genetic BP neural network.First,conduct grey correlation analysis on campus water data to find out the main factors affecting campus water use,and then input it into the genetic BP neural network model to first obtain the predicted value of the point,according to the residual value of the point forecasting and the actual value,Use this to estimate the upper and lower bounds of the forecasting interval.Interval forecasting can accurately predict the fluctuation range of water consumption in the future.It is compared with the forecasting results of BP neural network,Matlab is used to simulate the campus water interval data.The results show that the predicted data and the actual data are basically consistent The simulation accuracy of its forecasting data is 90.32%,which proves that the article method is feasible,and this forecasting method has certain reference significance.The campus water optimal scheduling model has many and complicated constraints and presents a non-linear problem,Aiming at the multi-objective scheduling problem of this model,the objective function is expressed by minimizing water cost,water environment and maximizing population target,a cuckoo-membrane algorithm is proposed.Membrane algorithms can solve the cuckoo algorithm,which is easy to fall into a local optimum when solving the campus water problem.Moreover,the algorithms in the membranes run independently.Various types of artificial intelligence algorithms can be added and used.The specific process is to divide the membrane structure into an auxiliary membrane and a main membrane,first,individual local optimization is performed in the auxiliary membrane,and then the optimized individuals are sent to the main membrane for global optimization.After an example analysis,the average water saving rate can reach 2.73%,which proves that this algorithm can achieve the purpose of optimizing the objective function.It is feasible to apply the cuckoo-membrane algorithm to campus water optimal scheduling and can be borrowed for other optimized scheduling.