Estimation of Spacecraft Telemetry Faults using SVM &K-Means Machine Learning Algorithms

Abstract

One of the most pressing issues in aeronautical engineering is the development of a makes an
effective satellite health monitoring. A spacecraft's vital systems must be able to withstand
the effects of the space environment. In order to avoid errors and handle important
circumstances, these systems are built to be complex. System status telemetry is transmitted
to a ground operator by the spacecraft, and the telemetry characteristics are used to monitor
the spacecraft's functioning. Using satellites telemetry data and integrated information,
modern machine learning and data mining technologies may create an advanced monitoring
system. In order to limit the danger of satellite failure, telemetry processing aids in the
visualization of satellite data. An overview of current data mining practices, as well as a
foundation of necessary processes for tackling telemetry data problems such as detecting
errors, prediction and summarization and visualization of enormous volumes, is presented in
this study.