Design and Implementation of the Fault Tolerance Module in PHOENIX CubeSat

摘要

The research in space technology and small satellite development has become more prevalent in recent years On-board flight software (FSW) should be able to handle subsystems with recovery capability in the presence of errors and faults Fault Detection Isolation and Recovery (FDIR) has become a key function when designing flight software This thesis depicts the FDIR functions that are implemented in the On-Board Computer (OBC) of PHOENIX For the QB50 mission the main objective of PHOENIX is to conduct research with respect to the lower atmosphere and to study the atmospheric re-entry process associated with aerothermodynamics phenomena According to this objective two science payloads the INMS and SolarEUV are installed in PHOENIX Instead of routine operation the ground station has to upload scripts to control these two payloads This increases the complexity of the on-board software Thus robustness and reliability become one of the main requirements of the FSW development FSW is in charge of command validation and execution mission scheduling data reception storage and downlinking and maintaining satellite health The OBC is also the interface for managing the communication between subsystems and payloads The cause of system failure can be very complex including software hardware and human factors Besides human factors many failures are caused by two primary effects of radiation: Single Event Upset (SEU) and Single Event Latch-up (SEL) SEU is a change of state in micro-electronic devices which may cause system crash behavior change and data damage SEL is a current caused by high-energy particle collision which can damage devices However some of these failures can be prevented with the right software design and we need to find solutions to handle the remaining failures PHOENIX CubeSat uses several FDIR methods including a software supervisor watch dog timers data redundancy and so on to scan important system parameters for potential failures and also to avoid data corruption A FDIR library is designed to support the ability of the software supervisor to handle potential failures Ground station is able to update this library by unloading scripts These FDIR methods and strategies follow some guidelines principles and a failure hierarchy in the engineering phase and are able to adjust to single events and other failure situations Herein some discussions and lesson learned are presented

獎項日期	2015 7月 22
原文	English
監督員	Jyh-Chin Juang (Supervisor)