Research On Key Technologies Of Online Water Quality Early Warning In Yangtze River Basin

The safety of water supply is the guarantee of people's life health.Nearly 80% of the water from the Yangtze River Basin.However,the waters in the Yangtze River Basin have deteriorated year by year with the rapid development of social economy.The drinking water security of the people in Jiangsu Province is in danger because of the frequent occurrence of sudden pollution incidents.In order to provide refined management services for the regional water resources management department to ensure water supply security,modern information technology is used in this paper to carry out research on key technologies for online water quality safety assessment and early warning in the Yangtze River Basin of Jiangsu Province.The research of this paper has important theoretical significance and engineering application value.The main work of this paper is as follows:(1)The principal component analysis(PCA)method and K-means clustering algorithm are used to analyze the overall water quality characteristics of the urban-rural water supply in the Yangtze River Basin in Jiangsu Province.The water quality parameters that have a significant impact on water quality changes are determined.The normal variation of water quality parameters is summarized.And the main pollution index of water quality is analyzed.(2)A LSTM-based water quality parameter change trend prediction model is proposed and good results are achieved.In order to improve the historical data set which was used to train and test the LSTM-based water quality parameter prediction network structure,a data filling algorithm is studied aiming at the problem of data loss in online monitoring time series.(3)A water quality safety evaluation method based on game theory-cloud model is proposed.Combining the game theory-cloud model evaluation method with the watershed characteristics research results,the safety evaluation method of the Yangtze River Basin in Jiangsu Province was constructed.The accuracy of water quality safety assessment is further improved by refining water quality assessment standards and incorporating time series features into the water quality safety assessment process.(4)In view of the problem of insufficient dynamic adaptability of the existing sudden water quality warning model,the online measured data is introduced to revise the early warning results in real time with Kalman filter.The self-optimization of two noise parameters Q and R by Kalman filtering is carried out by using chaotic particle swarm optimization.Finally,a realtime correction method for sudden water quality warning based on chaotic particle swarm optimization is constructed.Experiments show that the proposed method can effectively improve the prediction accuracy of pollutant diffusion and has strong dynamic adaptability.(5)Combining with the requirements of the subject,this paper makes a functional analysis of Jiangsu Province's urban-rural water supply supervision platform.The key technology research results are deployed on the platform containing the application of a series of functions,such as on-line monitoring of water quality,prediction of water quality trend,water quality safety assessment and early warning of sudden pollution incidents.