基于注意力机制的跨境断面水质预测模型研究  

Research on cross border section water quality prediction model based on attention mechanism

作　　者：朱齐亮 余雪婷 ZHU Qiliang;YU Xueting(School of Information Engineering,North China University of Water Resources and Electric Power,Zhengzhou 450046,China)

机构地区：[1]华北水利水电大学信息工程学院,河南郑州450046

出　　处：《现代电子技术》2024年第16期145-149,共5页Modern Electronics Technique

基　　金：国家自然科学基金面上项目:支持密文检索的单像素加密原理及方法研究(62275080)

摘　　要：为了充分掌握水体污染治理效果,为水环境保护和生态治理提供科学依据和技术支持,提出一种基于注意力机制的跨境断面水质预测模型。运用长短期记忆神经网络建立跨境断面水质预测模型,引入注意力机制,建立跨境断面水质预测序列编码矩阵。利用长短期记忆神经网络解码器对序列矩阵数据进行解码操作后,输出跨境断面水质的预测结果。实验结果表明,所提模型可有效提取跨境河流纵向断面水质化学需氧量(COD)时间特征与数据特征,同时可预测跨境纵向断面水质内的余氯、浊度等,且预测跨境断面水质高锰酸盐指数较为准确,应用效果较佳。In order to fully grasp the effectiveness of water pollution control and provide scientific basis and technical support for water environment protection and ecological governance,a cross-border section water quality prediction model based on attention mechanism is proposed.A cross-border section water quality prediction model is established by means of long short-term memory neural networks,and a sequence coding matrix for cross-border section water quality prediction is established by introducing attention mechanisms.After decoding the sequence matrix data using a long short-term memory neural network encoder,the predicted results of cross-border cross-sectional water quality are output.The experimental results show that the proposed model can effectively extract the temporal and data characteristics of COD(chemical oxygen demand)in the longitudinal section of cross-border river,and can also predict residual chlorine and turbidity in the water quality of cross-border longitudinal sections.The prediction of the permanganate index in the water quality of cross-border sections is more accurate,and the application effect is better.

关 键 词：注意力机制 长短期记忆神经网络 跨境断面 水质预测 序列编码矩阵 编解码器 化学需氧量(COD) 

分 类 号：TN911.23-34[电子电信—通信与信息系统] TP302[电子电信—信息与通信工程]