摘要
以RBF网络建立水质评价模型,网络的输入为某水样中参与评价的n种水质指标实测值的集合,输出为该水样的水质级别。问题关键集中在网络结构的选择、可调参数的优化方法,以及学习样本的代表性上。将n个样本划分成c个类的划分序列。只要在一定水平时样本被归入同一类后,在进行更高水平的划分时,确定类数即确定分类结果。
Use RBF network to establish water quality assessment model. Its input of network is aggregations of effective test results, which based on N kind of water quality samples gathered by water sample prepared to evaluate, and the output is the water samples＇ grading. Key problems are the choosing of network structure, adjustable parameter optimization method, and representation of the learning sample. The N samples need C category division, once the samples were classified into the same category under certain level, the categories-classified results would be realized even under a more detailed classification.
出处
《兵工自动化》
2008年第4期32-33,38,共3页
Ordnance Industry Automation
关键词
水质评价
人工神经网络
RBF网络模型
Water quality assessment
Artificial neural network （ANN）
RBF neural network model
