| Townlet is a political, economical and cultural centre to the rural area , it is not only tie and bridge between town and country, but also a significant base, that can propel the development of town and township enterprise, carry the science and technology, culture and education a step forward for rural area. However, In recent years, following the fast development of townlet, ecological environmental quality of townlet in our country is becoming terrible, and the development about the research on quality assessment is also becoming fast. Building a environmental quality evaluation indicator system (EQEIS) of townlet, establishment about the methods of weighting and the models of ecological environmental quality evaluation emerge at a historic moment. However, these hasn't been a standard environmental quality evaluation indicator system, and the methods of weighting and assessment have not been mature, therefore, the research on building an environmental quality evaluation indicator system for townlet was taken as the point of the department. Then, the methods of weighting and the models of ecological environmental quality evaluation joined in the discussion. All of these can provide scientific basis and reference for the eco- environmental management, eco- environmental engineering, eco- environmental programme and construction.In this paper, based on refering to the large number of domestic and foreign research results, the study on the methods of weighting and the models of ecological environmental quality evaluation were finished. The research results are as follows:(1) On the basis of a systematic analysis of the eco-environmental quality evaluation concept, the evaluation criteria and methods and the evaluation indices system, according to the target of environmental protecting programme in our country, the technology provisions of eco-environmental quality evaluation, and Chinese grazioso environment quality assessment index system for townlet, in the light of dushan's own eco-environmental characteristic, the environmental quality evaluation indicator system(EQEIS) of townlet was picked out from the preelection indicators by four methods (Frequency statistics method, Theory with practice method, Delphi method, Correlation analysis method) , EQEIS of Dushan town was buit in this research, which was consist of 3 aspects (natural source, regional environment, the level of regional economy) , 12 subentries and 28 single factors. That could express the primary aspects of the townlet's ecological environmental quality. And that could be used to evaluate the townlet's environmental quality objectively and systematically.(2) Based on Delphi-AHP, principal component analysis (PCA) and BP- Neural network, the weight of the indicator system arrived at a rational conclusion. It turned out that, Delphi-AHP is a simple and convenient process, it could be mastered easily. But, the algorithm is very coarse and subjective, so, the precision is uncertain. PCA is forthright, objective and comparable. However, in terms of theory, this conversion is linear, and the sample must be adequate and satisfy some specially appointed conditions. As a result that, sometimes, it could not agree with reality. BP- neural network is extracted regular characters of each sample from the training samples, then to recognizes test samples, which have some advantages of strong fault tolerance, self-learning and high accuracy. Comparison with three methods, the conclusion was that, we could choose Delphi-AHP as reference, choose BP- Neural network to determine the details of the weight.(3) In the basis of weight, AHP-comprehensive index model, PCA-cluster analysis model and BP- Neural network model were used in dushan eco-environmental quality assessment. Compare with three models, we found that, if the weight of the indicator system was determined, AHP-comprehensive index model could fit eco-environmental quality evaluation very well. PCA-cluster analysis model emphasis on indicator system's own objective inherent laws, it could be appropriate for eco-environmental quality subjective evaluation. When the sample is abundant, BP- Neural network model could be perfect for eco-environmental quality assessment. Although it is a detailed and complicated procedure for choosing and establishing the sample, once the model is set up, and can train successfully, it could combine conscious activity with objective reality very well. Not only the result could be accurate, but also the follow-up process could become easy. Therefore, the BP- Neural network model could be worth using and spreading.In this research, Dushan was chose for example. The analysis result revealed the law of spatial distribution of eco-environmental quality in Dushan area: from centre to four sides, the grade of eco-environmental quality decreases by degrees. Lukou village, as the centre of the area, whose eco-environmental quality is the most fragile, Xiawang and Xudong village come second. Donggang and Xiantai village are the average level. In the west of study area is Dushan and Fandun village, which are fine quality. And the eco-environmental quality of Keying village is the best. Then the study on the correlativity between eco-environmental quality and influent factors was done, it turned out that, the relevance between eco-environmental quality and social economy was maximum, the level of environment was second, natural resource was minimum. The conclusion is match with the fact, it proves that the indices system and evaluation model is practicality and reliability. |
PDF Full Text Request