This content is not published

Automatic Spacecraft Status Characterisation By Data Mining Mission History

Thu, 02/21/2019 - 18:56
Start Date: 
End Date: 
Programme Reference: 
SATE - Systems and Advanced Technologies Engineering

The main purpose of this project, as mentioned in the SoW, was to analyse historical mission data and search for new features which would be useful in the telemetry checking, command verification and procedure writing process in the future.
Features of a signal (or parameter in this context) are defined as characteristics that can be measured or calculated over a given time window (such as mean, standard deviation or minimum of a signal).
In the SoW, it is also defined that useful features in this context are those that are predictable enough under known circumstances that a check could be constructed which would not result in too many false positives.
It is also said that each parameter is expected to have some features that are useful and some that are not. In addition, some features are expected to be only useful under certain conditions, e.g. orbital effects, commands, or when other parameters are in certain states.
For this reason, the assessment of the usefulness of features was performed both over the entire time period (i.e. under no specific conditions) and over multiple time periods (i.e. under certain conditions). The data set used to extract the knowledge refers to one year of telemetry data of the VEX satellite (in the period 1 March 2013 - 28 February 2014). This analysis was conducted in WP1 and WP2 of the project.
Another goal of the project was to analyse global mission data, which have to be pre-processed and enhanced to create new data sets called derived parameter time series. These are intended to be used to identify the external factors that influence features checkability. The definition and extraction of derived parameters was performed in WP3.
The last step foreseen in the SoW was the identification of useful correlations between the derived parameters, the telemetry data and the various features.

Application Domain: 
Generic Technologies
Data treatment
Space applications
Executive summary: