Study On Water Quality Of Tap-water Based On Multi-Information Fusion Forecasting System

The development of national economy is closely related to the production of tap water,which lays a solid foundation for the stability and health of people's life.In recent years,the shortage of water resources is becoming more and more serious,the demand for water supply is increasing,the distribution of water sources is irregular and the raw water quality is complex and varied,which made the traditional surface water plant production face unprecedented challenges.At present,the feasible way to reduce and solve the impact of water pollution and improve the safety of water supply is to build water quality prediction system and safe water supply early warning system.Aiming at this problem,the experiment will carry out the relevant research on the construction of the tap water prediction system,and apply it to the water quality prediction of Tangshan city's waterworks.The main research contents and conclusions are as follows.First of all,the experiment monitors the incoming and outgoing water of the waterworks,collect the data of various indicators in tap water,such as p H,NTU(turbidity),oxygen consumption,number of bacteria,and record the weather conditions of the day.Then,the p H,NTU and oxygen consumption were divided into three levels: Level 1,level 2 and level 3,and PNN neural network was built according to the divided results and monitoring data.In the experiment,65 groups of data were selected from a large number of monitoring data,among which 51 groups were used as training samples of neural network,and the remaining 14 groups were used as test samples of neural network.MATLAB was used to simulate and test the grade,so as to predict the grade of water quality in waterworks.Finally,the PNN neural network and BP neural network were compared.The results show that PNN neural network model is more accurate and faster than BP neural network model in water quality prediction.Figure 24;Table 9;Reference 51...