Identifying High Area-to-Mass Ratio Space Objects Using Historical Two-Line Element Set
Description
The HAMRTLESeq series of datasets (in WEKA *.arff format) were constructed using RSO's TLE historical data for the time series data set of the HAMR classification. They are used to validate the two hypotheses presented in the paper "Identifying High Area-to-Mass Ratio Space Objects Using Historical Two-Line Element Set, 2019" (authors: Bin Liu, Junfeng Wu, Li Yao.) and forming a machine learning models for HAMR objects recognition.
Files
Steps to reproduce
The TLE historical data published by the USSTRATCOM catalog records the evolution of the spatial target orbit. The Space Target Catalog of the North American Air Defense Command (NORAD) contains tens of thousands of HAMR objects and other types of spatial targets. Therefore, these two data sets can The orbital time series data set used to construct the classification of HAMR objects. This article first obtained NORAD IDs for 156 spatial targets from the NORAD catalog, including 80 HAMR objects and 76 other types of RSOs. Then, download the TLE historical data file from 1958.01.01 to 2014.09.14 from the USSTRATCOM catalog according to the NORAD ID. Then, six orbital parameters and two features are extracted from each TLE historical data file to form a historical feature data set Feature of the corresponding spatial target. Then, from each feature set, 2, 4, 8, and 16 records with a time interval of 24±0.3 hours are recorded to form a feature sequence, and a total of 137 spatial target feature sequences are obtained, which are 137 Space targets include 64 HAMR objects and 73 other types of RSOs. Finally, based on these feature sequences, a series of time series data sets consisting of 2, 4, 8, and 16 observations of TLE feature data in 1, 3, 7, and 15 days, including HAMRTLESeq_1day, HAMRTLESeq_3days, HAMRTLESeq_7days, and HAMRTLESeq_15days, are collectively referred to. It is a HAMRTLESeq series time series data set; at the same time, a data set composed of one observation TLE feature data is formed, named HAMRTLE. These data sets will be published on the Mendeley Data website (key words: HAMR TLE), as shown in Table 2 for each data set.