学习率有限监督调整BP网络在水质评价中的应用

以地表水环境质量标准基本项目标准限值为依据,随机生成标准样本和检测样本,采用学习率有限监督调整BP网络建立水质评价模型。完成网络训练的模型对检测样本进行检验评价,检验结果表明,该人工神经网络水质评价模型具有较高的精度,同时,可以避免人为主观因素对水质评价的影响,保证了水质鉴定的科学性和公正性。

Application of BP Networks Based on Limited Supervising to Learning Rate Regulating to Water Quality Evaluating

In the research of this paper,based on the surface water environmental quality grade scale,standard samples and test samples are created randomly.BP network with self-adapting learning-factor,named limited supervising to learning rate regulating,is applied to train the water quality evaluating model.By the trained networks,the evaluated results based on the test samples,show that the ANN model can evaluate water quality with high precision,at the same time,the method can evaluate water quality objectively and justly,not subjectively.