摘要
采用改进的动量-自适应学习率调整的 BP 神经网络,以国家水环境质量标准作为训练样本,以黄河流域的大汶河水系的水质监测数据作为检验数据建立了一个多前馈网络数学模型,应用此模型对浑河沈阳区段的水质进行评价,由于网络的改进有效地提高了训练的速度、节省训练时间,并且提高了评价的准确性。
The article applies neural network to estimate the water quality of Hunhe. Pollution indices are adopted as training example to train the network and observation data of The Yellow River are used as testing data Because of the improvement of network, the speed of training is improved and accuracy is also increased.
出处
《辽宁工程技术大学学报（自然科学版）》
CAS
北大核心
2004年第5期616-617,共2页
Journal of Liaoning Technical University (Natural Science)
关键词
改进
BP神经网络
水质评价
浑河
improve
BP neural network
water quality assessment
hunhe river
