Study On The Water Quality Assessment Of The Harbin Section Of Songhua River Based On Principal Componet Analysis And BP Neural Network

With the rapid social and economic development, to the natural human endless,wanton destruction of our home, bear the brunt of the problem of water environment. It is very urgent to improve water environment quality. To evaluate the water quality can be learned that the current situation of water environment quality,provide scientific basis for the management of water environment. As one of the seven main river,Songhua river carrying around life and production. This article spreads out its researeh on water quality of river basin to Songhua river,according to the specific condition of Harbin Songhua river section, in the comprehensive consideration o f all kinds of natural and human factors affecting water quality of Songhua river,on the basis of combining the materiality principle this paper selected the includingCOD, BOD5 and NH3-N,18 index as the basis of the Songhua river water quality evaluation index in water monitoring.Through principal component analysis to research index,finally chose COD,BOD5,CODMn,NH3-N,TP,and oil such as BP neural network to evaluate water quality of water quality indicators.The study data for water quality assessment is monthly monitoring indicators of 6 sections include Zhushuntun Ashihekounei Ashihekouxia Hulanhekounei Hulanhekouxia Dadingzishan from 2009-2013.Evaluation Water quality according to the years and for during dry seasons,common seasons and wet seasons.Water quality evaluation method with principal component analysis and BP neural-network.Through principal component analysis principal component values and the order of pollution of each section,provide indicators based on BP neural network to simulate water quality.By training the BP neural network to the optimal evaluation model,through the model to evaluate each time section respectively, for each section of the pollution level, water quality classification of each section.In this paper,principal component analysis to choose software for SPSS22.0,BP neural network is chosen MATLAB2010 a.Evaluation results show that the water quality in six sections of the worst for Ashihekounei section within five years ranked in the last one,the water quality areⅤof water quality,the best for Zhushuntun section,Ⅱ- Ⅲ for water quality,the remaining four section water quality is good,achieves the planning of the water quality requirements.Through two kinds of methods to evaluate water quality situation of Harbin Songhua river segment,which will provide a scientific basis for improving water quality and governance.