Computer modelling of Spacecraft Performance for Flight Operations:
Lessons Learned taking a model from Concept to Operational
Software


Richard Dragonette
Richard-Dragonette@jhuapl.edu
JHU
Abstract:
A large satellite is a complex problem from the perspective of
modelling its thermal response, and the response of its power distribution
subsystem to strenuous data collection experiments (DCE's).  If it
desired to execute a number of DCE's closely spaced in time, a model
which predicts the spacecraft's response during data collection, as well
as during its recovery, is required to ensure the satellite can execute the
desired task and be satisfactorily recovered prior to the start of the next
DCE.

The author has been, and is currently, involved in many phases of the
development of such an automated software model for a large earth observing
satellite.  Areas of extensive involvement include:
developing some of the analog models implemented by the software; writmg
software specifications; redefining the scope of the model; writing the
software; testing and verifying the software; and developing  the
verification criteria for the model.

This paper presents some of the positive and negative lessons learned
during the evolution of the software model. Topics of discussion include:
defining what the model needs to do versus what it could do given infinite
resources; the need for simplification even with powerful computing tools;
short term single DCE simulation versus longer term simulation;
setting definitive goals upon which to judge the completeness and
accuracy of the model; and on-orbit model parameter modification.