Research And Application Of Water Quality Prediction Method Based On EEMD-LSTM

Water is an important resource for human survival.It plays an extremely important role in all aspects of human production.At the same time,it is also a basic organic part of ecological environment construction.It is irreplaceable by any other resource.With the great success of China’s economy in recent years,Chinese industrialization level has increased significantly,but at the same time,surface water is often polluted by industrials.Water quality prediction has been a basic work to protect surface water resources and an essential means to deal with water resources climacteric.Water quality prediction reflects the trending of surface water quality in the future accurately,which is of great significance for enhancing the conservation and utilization of water sources,improvising the current situation of protection,and promoting the restoration of the ecological environment.With the swift advancement of artificial intelligence,data analysis and prediction methods based on deep learning have provided a new idea for water quality prediction.Aiming to boost the accuracy and generalization ability of the water quality prediction model,based on the actual water quality data of the Taihu Lake and historical meteo-rological data,this dissertation proposed a method combining the Ensemble Empirical Mode Decomposition method and the Long and Short-Term Memory neural network to construct EEMD-LSTM water quality prediction model.The main research work is demonstrated as follows:The study and realization of a multi-feature water quality prediction model based on EEMD-LSTM.Considering the dual factors of hydrology and meteorology,this dis-sertation proposed a multi-feature water quality prediction model based on water quality and meteorological features.Compared with a single feature model that only considers a single water quality feature,the accuracy of water quality prediction is improved.At the same time,this dissertation used the EEMD algorithm to decompose the time series into several sub-sequences,amplifies the details in the time series data,which makes the fluctuation degree of the sub-sequences more stable than the original series.Therefore,the proposed method solved the prediction lag problem of the LSTM network.The programming and implement of the intelligent water quality prediction sys-tem.The EEMD-LSTM prediction model proposed in this dissertation is employed on the above mentioned prediction system.The system adopts the B/S distributed archi-tecture design,using HTTP to completes the front-end data communication interaction and water quality prediction model deployment base on Django,and performs stress testing on the system to show the stability of the system.The intelligent water qual-ity system finally realized has user management,water quality data management,and models.Management and water quality prediction functions.