Research And Evaluation On The Quality Of Water In Jian Jiang River

At present, the index methods are often used in water quality evaluation, although the calculation is simple and easy to get the results, the results are not representative,and it’s hard to compare in horizontal and vertical, also, these methods can’t study the relationship between internal factors and trace the source of pollution.Based on this, this paper focuses on nine monitoring stations in the city district of Jian Jiang river in DuYun. Firstly review how the Jian Jiang play a important role in the economic development and people’s life of DuYun, and then, three methods, namely fuzzy comprehensive evaluation, principal component analysis and the BP neural network toolbox of MATLAB, are chosen to evaluate the environment quality in nine monitoring stations.As a comparative object, the single factor index method was firstly used in water quality evaluation of Jian Jiang river, the result is not ideal, for the other three methods,although the evaluation mechanism is different, the three methods are considering the relative impact of each factor in the water environment, so the evaluation result is superior to the single factor index method. The results showed that the nine monitoring stations are evaluated for two level V and two level IV, and the rest are level III. Besides,there are too much BOD5 and COD which are the main factors affecting the water quality because of the higher organics in the water environment.Wherein, when the fuzzy evaluation method and BP neural network are used for water quality evaluation, the classification results can be established directly. The former are somewhat complicated in processing the data, but the latter, which with a sample structure, can be integrated in the MATLAB. As for the principal component analysis, which tend to analyze the internal correlation in factors, it can get the relative influence of factors for the water environment by using SPSS statistical software, at the same time, the main factors affecting water quality and the sort of monitoring stations can be established according to the total variance explained and loading matrix.In view of the three methods have their own advantages and disadvantages,combined with the standards of water quality evaluation method should be simple,efficient and practical, finally determined the principal component analysis and BP neural network to be a comprehensive evaluation method, among them, using BP neuralnetwork to predict and classify water quality, and the principal component analysis method is used to analyze and evaluate the relative effect of factors in the water environment, and then determine the main pollution factor and make management countermeasure, so as to get quantitative description and qualitative classification for the water environment.