摘要
当前水文自动测报过程依靠无线通信技术进行数据传输,对地面环境因素的干扰较敏感,最终显示的测报结果平均绝对误差(mean absolute edrror,MAE)较高。因此,提出基于北斗卫星通信技术的水文自动测报方法。基于信息熵确定遥测站点布设方案,实现自动数据采集。结合联邦学习架构和长短时记忆网络的生成对抗网络模型,识别接收数据异常信息。根据遥测数据确定当前水位,生成可视化水文自动测报结果。实验结果表明:新研究方法的水文测报结果MAE值保持在0.15以下,显著提高了测报精度。
The current automatic hydrological measurement and reporting process relies on wireless communication technology for data transmission,which is sensitive to interference from ground environmental factors.The final displayed measurement and reporting results have a high mean absolute error(MAE).Therefore,a hydrological automatic measurement and reporting method based on Beidou satellite communication technology is proposed.The telemetry station layout is determined based on information entropy to realize automatic data acquisition.A generative adversarial network model combining federated learning architecture and long short-term memory network is used to identify abnormal information in received data.Current water levels are determined based on telemetry data,and visual hydrologic automated measurements and reporting results are generated.The experimental results show that the MAE value of the hydrological measurement results of the new research method remains below 0.15,significantly improving the accuracy of the measurement.
作者
王巧丽
鲁青
WANG Qiaoli;LU Qin(Bureau of Hydrology,Changjiang Water Resources Commission,MWR,Wuhan 430000,China)
出处
《自动化技术与应用》
2025年第11期135-140,180,共7页
Techniques of Automation and Applications
基金
国家重点研发计划项目(2022YFC3200129)。
关键词
北斗卫星通信
遥测站点
数据传输
异常修复
水位
水文测报
Beidou satellite communication
telemetry station
data transmission
abnormal repair
water level
hydrological measurement and reporting
