基于BP人工神经网络的高锰酸盐指数预测模型

CODMn forecasting model based on BP ANN

高锰酸盐指数(CODMn)作为有机物相对含量的指标之一,同时是河流水质规划的一项重要指标。CODMn预测过程可认为是一种复杂非线性函数关系的逼近过程,利用BP人工神经网络建立CODMn预测模型,将历史样本作为模型的输入进行训练,利用训练好的网络模型进行预测,结果表明该模型具有良好的预测性能。

CODMn can be used as the indicator of the relative content of organic substance, and it is an important index of river water quality plan. CODMn prediction process can be considered as a complex nonlinear function approximation process, so this paper employs BP artificial neural network to construct CODMn forecasting model. It therefore can train such BP ANN with the collected data before CODMn forecasting with the well-trained neural network. The experimental results show that this approach has better forecasting capability.