Spacecraft Data Anomaly Analysis Based On Large Data

Along with the increasing investment in the field of spaceflight,the number of on-orbit spacecrafts has shown a sharp increase pattern.With the increasing requirements of control accuracy,working mode and data complexity of on-orbit spacecrafts,the management and maintenance of spacecrafts and the handling of anomalies have brought tremendous work.Challenges.From the ground design of spacecraft to launch and then to on-orbit operation,it will be affected by various complex conditions,resulting in varying degrees of anomalies.With the increase of time and the number of spacecraft launches,a huge amount of data has been accumulated,which provides us with a new way to solve this problem: applying data mining related to large data fields to analyze the abnormal problems of spacecraft.But because the factors causing the anomalies are too complex,if all the research is carried out,the effect may not be good.Therefore,this paper mainly uses a small part of the large amount of data accumulated in the relevant literature to carry out the experiment.Firstly,the attributes in the abnormal database are filtered to find the abnormal attributes with high cohesion,and the attributes are sliced,sliced,and kMeans clustering.Then,the attributes are coded to construct the data model which conforms to the data characteristics and the application of the mining platform.Secondly,optimize the data mining algorithm,verify the effectiveness of the optimization through samples,and test the optimized algorithm through Weka data mining platform.In the process of the experiment,the rationality of the threshold of measurement parameters is evaluated with domain knowledge,and the attribute association rules with high reliability and cohesion are obtained.The anomaly analysis model is validated and evaluated.Finally,when similar anomalies are encountered again by anomaly analysis model,the causes of anomalies can be quickly analyzed,or when several high cohesion attributes occur at the same time,the possible anomalies can be handled in advance.