摘要
为支撑导航卫星星间链路网络的稳定运行,提升对星间链路网络的监控管理能力,解决地面系统,对星间链路网络状态感知迟滞等监视能力不足问题,以及对未来网络趋势的预测能力不足问题,提出并构建了星间链路态势感知模型,包括数据预处理、指标计算和态势预测三部分;对于监视型网络状态指标,通过对原始遥测信息进行抽取获得;对于更为复杂的分析型网络状态指标,首先利用人工规则进行数据筛选,随后利用数据挖掘算法提取有效特征,选取相关性较大的参数综合得出复杂指标;在此基础上,利用BP神经网络对星间链路网络态势进行预测;通过实验,某状态指标预测的准确率稳定在98%左右,结果表明该模型可以有效的对星间链路网络的未来状态进行感知预测,可以对星间链路网络状态异常进行及时处理,具有较高的可行性。
In order to support the stable operation of the inter-satellite link network of navigation satellites,improve the capabilities of monitoring and management in the inter-satellite link network,solve the insufficient problem of monitoring capabilities of the ground system in the inter-satellite link network state,such as the hysteresis,and predict the insufficient problem of capability in the future network trend,an inter-satellite link situational awareness model is proposed and constructed,including data preprocessing,index calculation and situation prediction;The indicators of surveillance network state are obtained by extracting the original telemetry information;The more complicated indicators of analytical network status,manual rules are first used to filter the data,then the algorithms of data mining are used to extract effective features,and the parameters with greater correlation are selected to synthesize complex indicators;On the basis,the inter-satellite link network situation is forecasted by using the BP neural network;Through the experiments,the prediction accuracy of state index is stable at about 98%,and the results show that,the model can effectively perceive and predict the future state of the inter-satellite link network,and can detect abnormal state of the inter-satellite link network,which is highly feasible to carry out timely processing.
作者
王海学
孙剑伟
田露
WANG Haixue;SUN Jianwei;TIAN Lu(th Research Institute of China Electronics Technology Group,Beijing 100083,China)
出处
《计算机测量与控制》
2022年第10期262-267,共6页
Computer Measurement &Control
基金
国防预研项目(GFZX03010105280203)。
关键词
星间链路
网络态势感知
数据挖掘
神经网络
遥测数据处理
inter-satellite link
network situational awareness
data mining
neural network
telemetry data processing