Research And Design On The Water Quality Data Fusion Platform

In recent years,with the increasingly serious pollution of water environment,many governments have established many water quality monitoring systems,but the data of each platform is independent of each other,and there are data islands,which are not conducive to the comprehensive analysis and utilization of data.In response to this situation,a water quality data fusion platform is designed using advanced technologies such as data mining technology,big data technology,cloud computing technology and database technology to integrate and manage water quality data,which comes from complex multi-sources,to realize intelligent learning of big data in water environment,statistical analysis and forecasting.It can help with the water environment department to carry out efficient and accurate water quality prediction.It can also provide strong support for the development of effective and reasonable water environment protection including the protection measures.The main research contents of this thesis include the following aspects:(1)Designing and implementing a water quality data fusion platform based on water environment data collection,processing integration,data storage,data mining and analysis,and data visualization process.The platform contains a series of functions such as water environment topics,multi-dimensional analysis,pollution source files,reports center,data management,and data analysis.The data collection method and data storage method of the platform are introduced.When the data is incremented to a certain amount in the later stage,the historical data will be migrated to Hadoop,and will be processed and integrated by using the ETL(Extraction Transformation Load,ETL)tool kettle.A new water quality multi-factor prediction model is designed for water environment monitoring data to provide effective data support for water environment management decision-making.Taking a certain surface water monitoring site in Guiyang Guizhou as an example,data collection andintegration is carried out through a water quality data fusion platform,and the water quality prediction model is used to predict the surface water quality.At the same time,this thesis also contains the design of the water quality data fusion platform,requirements analysis,interface design,database design and etc.These will be described in detail.(2)A water quality multi-factor prediction model based on LSTM(Long Short-Term Memory neural network,LSTM)neural network is designed.Taking the monitoring water pollution data factor as an example to establish a prediction model,the maximum and minimum normalization method is used to preprocess the water quality data of the monitoring site,simplifying the fluctuation and complexity of the data,and then it calculates the high dimension by K-Similarity noise reduction method.The cosine similarity of the vector in space is used to remove the noise.Finally,the LSTM is used for prediction,and the Adam algorithm is used to optimize to update the weight parameters of the model to reduce the loss.The prediction results are compared with the BP neural network,RNN and traditional LSTM neural network model,and the prediction of the new model is more accurate.(3)System implementation and functional testing.The function module code is implemented and deployed in a test environment that meets the requirements.The test cases are written for rigorous testing and analysis in order to make functions,performance and other aspects gradually meet the needs of users.The water quality data fusion platform improves the efficiency and quality of water quality data exchange and integration,and realizes the data sharing of water environment across departments and regions.At the same time,the analysis and prediction of water quality big data is realized,which is of great significance to the environmental protection industry.