The Applicability Of Data Mining Technology For Environmental Protection Management System

Environmental protection is one important guarantee of realizing the sustainable development for the whole country. As a basic state policy, it is pay more and more attention and supported by government and the public. Pollution source is one of the most important factors on destroying the environment, making influence of the sustainable development of the country.It is an important management object to the protection of the environment. Implementation of on-line monitoring of key pollution sources is one important means to realize the long-term management of pollution sources. Through online monitoring of the key pollution sources, it can improve the efficiency of the pollution management, master the dynamic information sources in time, and be helpful for supervision and finding enterprise environment illegal behavior. It can enhance the government's environmental management ability, grasp the key river water quality conditions, strengthen the river basin water environment management, ensure the drinking water safety, and ensure the safety and health of people's life by carrying out the surface water quality automatic monitoring system.The environmental protection management system composed by on-line monitoring of key pollution sources system, surface water quality automatic monitoring system and the other operation systems is important on building a safe and comfortable life and business environment of the public. Along with construction and operation of the environmental protection management system, the environmental management level can be improved onto a new stage. It is of practical meaning for improving the environmental management science and technology,and also the quality of the environment.Since 2008, the environmental protection management system has been constructed by Zhongshan environmental monitoring station. As one important part, the online monitoring system of key pollution sources has been collecting the emission data of key pollution sources in Zhongshan city. In the meantime, in early 2011, water quality automatic monitoring system was started to use in monitoring eight major sources of drinking water, which obtained the real-time data of surface water quality. All of the data was integrated in the environmental protection management system. How to make good use of the lager number of the data is the main research topic of this paper.Using data mining technology, these data of surface water quality has a very important significance on analysis and forecast of water quality for the city. In this paper, we mainly use the automatic monitoring of data of Xiaolan waterway, Madafen monitoring station and 7 key pollution sources upstream and downstream as the research object, conducting data mining technology, to carry out screen, pretreatment and attribute reduction. Then the BP neural network technology is used on model building to forecast COD_Mn, which is the key index of water quality of Xiaolan waterway.In this paper, the basic concept of data mining and its tools and techniques are introduced first. Then the data collection source of key pollution sources monitoring system and surface water quality on-line automatic monitoring system of the environmental protection management system are presented. It is also discussed that how to screen the biggest key factors for COD_Mn from different water quality monitoring indexes by simple Bayesian classification. In this paper it is introduced that how to apply the method of rough set on monitoring data attribute reduction, and use BP neural network techniques to predict COD_Mn of the next day of Xiaolan waterway. Finally, in the means of the exact figure and predicting error, comparison is carried out between this method and other prediction technology. It can come to the conclusion that the forecasts effect based on rough set of screening BP neural network is better than other prediction technology.Because the surface water quality automatic monitoring system of Zhongshan runs relatively short time, the data used in this article is still limited; this makes the prediction effect of a slight deviation. But with the system further perfecting and monitoring data accumulating, the training of the model in this paper can slowly increased, which can enhance model prediction accuracy. As the surface of the water quality model will be further perfected, it can provide important means for environmental monitoring station to Zhongshan and environment management department's decision. at the same time, it can improve rapid response capability of environmental monitoring for the pollution and pollution accident hidden danger.