摘要
为更加合理地利用水资源,保护水资源,对供水系统进行优化调度以及对用水量预测方法进行研究十分必要。本文提出了基于改进BP算法的用水量预测模型:首先介绍了BP神经网络的原理;然后设计了三层BP神经网络,利用改进的BP算法建立用水量预测模型,其自适应调整学习率,结合LM算法,引入动量因子,利用自主设计的训练函数对设计好的网络进行训练;最后对模型预测结果进行分析,并与标准BP算法进行对比。通过实验发现,该模型可对城市居民的日用水量进行合理预测,预测精度可达到相关要求。
In order to make more rational utilization of water resources and protect water resources,it is necessary to optimize the operation of water supply system and study the forecast method of water consumption. In this paper,a water consumption forecast model based on improved BP algorithm is proposed. Firstly,the principle of BP neural network is introduced. Then,a three-layer BP neural network is designed. The improved BP algorithm is used to establish the forecast model of water consumption,which makes self-adaptive adjustments of learning rate,combines with LM algorithm,introduces momentum factor and uses the self-designed training function to train the designed network. Finally,the forecast results of the model are analyzed and compared with the standard BP algorithm. Experiments show that the model can reasonably forecast the daily water consumption of urban residents,and the forecast accuracy can meet relevant requirements.
作者
袁玉英
罗永刚
孙立云
YUAN Yuying;LUO Yonggang;SUN Liyun(School of Computer Science and Technology,Shandong University of Technology,Zibo 255000,China;School of Electrical and Electronic Engineering,Shandong University of Technology,Zibo 255000,China)
出处
《水资源开发与管理》
2022年第9期28-33,共6页
Water Resources Development and Management
关键词
用水量预测
模型
改进的BP算法
water consumption forecast
model
improved BP algorithm
