Research On The Application Of Support Vector Machine In Water Quality Monitoring Information Fusion And Assessment

Water pollution, which imperils the development and living conditions of human beings, has been a main subject in environment protection realm. The water quality monitoring and water quality assessment are one of the most important ways to manage and control the quality of water resources. As a first-phase work of water-quality improvement and water-resources management, they provide a practical scheme and scientific basis for the protection and integrated exploitation.Based on Matlab, by analyzing the data monitored by multi-sensor on the ground and the multi-spectrum remote sensing image, and from information infusion angle, Support Vector Machine methods is first introduced to the process of water quality monitoring and assessment, and their feasibility and validity are analyzed. The main contents and conclusions are as follows.Firstly, SVM model for water state recognition is designed and established based on ground monitoring data. Through analyzing the water state of Changjiang River and comparing with single-factor method, BP neural network method and D-S theory method, SVM shows its feasibility and better classification capabilities of water state recognition on water quality monitoring and assessment.Secondly, considering the monitoring of monthly water parameter average chroma as a time series prediction program, SVM model for water monitoring information fusion processing is established based on ground monitoring date. The designed model is applied on the monthly high potassium permanganate index monitoring of Lake Tai. And compared with RBF neural network model, the result show the SVM model is feasible to applyed on water quality monitoring, and provides a proven method for indirect water quality monitoring.Thirdly, Water quality monitoring model and water quality assessment model based on the remote sensing technology are respectively built based on the association of remote sensing image and ground monitoring iformation. Using gray degree values picked out from remote sensing image to inverse the concentrations of ground water parameters and water grades, the remote sensing imagery data is made full use to complements ground survey data, so as to obtain more water quality information, and have a better water quality assessment effect.