| 灰色聚类法和人工神经网络在水质综合评价中的比较 |
| |
| 引用本文： | 尹姗姗,朱雅敏.灰色聚类法和人工神经网络在水质综合评价中的比较[J].绿色科技,2014(3):75-77. |
| |
| 作者姓名： | 尹姗姗 朱雅敏 |
| |
| 作者单位： | 国家海洋局珠海海洋环境监测中心站; |
| |
| 摘 要： | 探讨了两种水质综合评价方法：改进的灰色聚类法和人工神经网络法。通过采用增加训练样本和黄金分割的隐合层节点优化算法建立了人工神经网络模型，将两种水质综合评价方法进行了比较，结果表明：改进的灰色聚类法计算量较大，主观性较强，评价结果稳定。BP人工神经网络进行水质综合评价具有客观性，但网络训练较为繁琐，通过插值生成训练样本，极大地增强了网络的稳定性。但扩充后的训练样本，不能代表复杂的水质实况，使评价结果受到一定影响。
|
| 关 键 词： | 灰色聚类 人工神经网络 水质 评价 |
Comprehensive Assessment of Water Quality by Using Gray--clustering Law and Artificial Neural Network |
| |
| Yin Shanshan,Zhu Yamin.Comprehensive Assessment of Water Quality by Using Gray--clustering Law and Artificial Neural Network[J].LVSE DASHIJIU,2014(3):75-77. |
| |
| Authors: | Yin Shanshan Zhu Yamin |
| |
| Institution: | (State Oceanic Marine Environmental Monitoring Center of Zhuhai ,the State Oceanic Administration of China ,Zhuhai 5190152 ,China) |
| |
| Abstract: | This article discusses two kinds ofwater comprehensiveevaluation methods：the modifiedgray--clustering law and the artificial neural network. The artificial neural network model is created through increasing the number of training samples and optimization algorithm of golden section hidden layer node. Comparing the two kinds of water comprehensive evaluation methods, the results show that the modified grey--clustering method is rather stable and subjective,but its calculation burden is too heavy;the artificial neural net work is objective and more stable by interpolating, while the network training is tedious and the expanded training sample can not representscomplex water quality,so the assessment result will be influenced. |
| |
| Keywords: | gray-- clustering law artificial neural network water quality assessment |
| 本文献已被 CNKI 维普 等数据库收录！ |
