首页|一种卫星遥测数据多参数预测方法

一种卫星遥测数据多参数预测方法

扫码查看
针对目前遥测数据多参数预测精度不足的问题,文章提出一种基于图注意力网络和时域卷积网络的预测方法.首先,采用多尺度时域卷积残差网络组件提取遥测时序数据在不同时间跨度下的时间依赖关系,以捕捉时间模式;随后,利用图结构学习组件自动获取遥测数据变量之间的空间依赖关系,以捕捉空间模式;最后,将图节点特征表示与数据嵌入表示进行融合,增强图注意力网络在信息聚合和消息传递过程中的学习能力.在某卫星遥测数据集上的应用结果表明:该方法比双向长短期记忆网络(LSTM)模型的平均绝对误差(MAE)降低 62.38%,显著提高了遥测数据多参数的预测精度,为保障在轨卫星正常运行提供了更多决策分析支持.
A Multi-parameter Prediction Method for Satellite Telemetry Data
In response to the current problem of insufficient accuracy in multi-parameter predic-tion of telemetry data,a prediction method based on graph attention networks and time-domain convolutional networks is proposed in the paper.Firstly,a multi-scale time-domain convolutional residual network component is used to extract the time dependencies of telemetry temporal data under different time spans,in order to capture temporal patterns.Subsequently,a graph struc-ture learning component is used to automatically acquire spatial dependencies between telemetry data variables,in order to capture spatial patterns.Finally,the graph node feature representation is fused with the data embedding representation to enhance the learning capability of the graph at-tention networks in the process of information aggregation and message transmission.The appli-cation results on a certain satellite telemetry dataset show that the method reduces the MAE by 62.38%compared to the LSTM model,significantly improving the prediction accuracy of multi-ple parameters in the telemetry data,and providing more decision-making and analysis support for guaranteeing the normal operation of satellites in orbit.

satellite telemetry parameterprediction modelgraph attention networktime-domain convolutional network

林启杨、张昊鹏、皮德常

展开 >

南京航空航天大学,南京 211106

北京空间飞行器总体设计部,北京 100094

卫星遥测参数 预测模型 图注意力网络 时域卷积网络

2024

航天器工程
中国空间技术研究院总体部(北京空间飞行器总体设计部)

航天器工程

CSTPCD北大核心
影响因子:0.552
ISSN:1673-8748
年,卷(期):2024.33(3)