Ecological Planning For Small Town Based On BP Artificial Neural Networks

Urbanization is a comprehensive issue, setting economic, social and environmental problems in one. Its essence is to optimize rural industrial structure and allocation of resources continually. The development of small towns is different from the one of metropolis. Small towns are in between artificial and natural environment, have close relationships with agro ecosystem and dispersed population. That is benefit, on the one hand, to maintain a good ecological environment of small towns by the conversion and degradation of pollutants. On the other hand, in the process of construction of small towns, if lacking reasonable plans, the unpredictable damage of the ecological environment will occur. In the process of traditional ecological planning, the accuracy and operability of planning are declined by the excessive anthropogenic interference, which causes an unsound guidance of the ecological construction of small towns. Thus, in this paper, artificial neural network(ANN), a computational model, was introduced into ecological planning. A ecological suitability evaluation model was developed by using (Back-Propagation)ANN and used in practical assessment.In this paper, based on the previous studies, the level of ecological health in Zhonghe town, Meihekou city, was evaluated firstly. The result showed that the ecological status of Zhonghe town was at the level of sub health. And then, with the help of MATLAB7.0, a suitability evaluation model, its data obtained by actual investigation, was developed based on ANN model and ecological planning theory. Finally, comprehensively considering the result of ecological suitability evaluation and the level of ecological health, ecological regional division program of Zhonghe town was brought forward. The feasibility of using BP ANN model in ecological planning was demonstrated by the result of research. Moreover, this new evaluation method have the advantages of simple ness, objectivity and high accuracy.