基于B-P人工神经网络的环境测点的优选

为了对有多项污染物的环境测点进行优选,提出了单项污染物对环境作用的“相对贡献率"和多项污染物对环境综合作用的“作用和贡献率"的新概念及其计算公式。将B-P神经网络原理与逐步聚类分析思想相结合,用于环境测点聚类优选。该方法用于成都市12个环境测点的优选结果符合客观实际。此外,它还具有简便实用、客观性好的特点。 

Optimum Setting of Environmental Monitoring SitesBased on B-P Artificial Neural Network

In order to optimize environmental monitoring sites with multi pollutants,the new concepts of relative contribution rate of single pollutant and accumulative action contribution rate of multi pollutants for the environment are proposed respectively, with the formula presented for the calculation of accumulative action contribution rate.The environment monitoring sites are optimized by the combination of B-P neural network principle with stepwise cluster analysis thoughts.The optimization results of 12 environmental monitoring sites of Chengdu tally with the demands,and the method is characterized by its simplicity,practicality and good objectivity.