Design and implementation of intelligent decision
making system to generate flight scenario
automatically
Seyyed Mehdi Dehghan
Researcher
Space Research Institute (SRI)
Tehran, Iran
smm.dehghan@gmail.com
Seyyed Hadi Cheheltani
Researcher
Space Research Institute (SRI)
Tehran, Iran
hadicheheltani@gmail.com
Ali Kiapasha
Researcher
Space Research Institute (SRI)
Tehran, Iran
Abstract— in this paper, a new structure of an autonomous
system generating a valid flight scenario for a satellite is designed
and implemented. The hypothetical mission of this satellite is
imaging from any points which are specified by users and also
acquiring a sequence of imaged from high priority points on
earth like fires jungles or volcano. The main intelligent decision
making system in the satellite has to generate flight scenario of a
microsatellite by receiving longitude and latitude of each
requested imaging point from the earth. In this article an
introduction of intelligent systems and their architecture for a
typical microsatellite, cooperation scenarios between ground
stations and a satellite and their missions are investigated. Finally
the proposed intelligent system architecture and implementation
of each layer and their integration is discussed. To test this new
developed architecture a proper test bench is designed too. One
of the approaches is adding some degree of autonomy to OBC of
a satellite. All goals that are mentioned for an intelligent system
and expansion in autonomy lead to self-based decision making
process. This onboard decision making process is helpful in those
situations that are not previewed on earth therefore are not
scheduled to be done by satellite. This facility helps the system to
pass the unsafe situations and response to them in a best way. At
least this aspect can reduce destructible side effects of sudden
situations.
Keywords- satellite autonomy, intelligent sysetm, planner-
scheduler, software in the loop, flight scenario
I. INTRODUCTION
The greatest motives for future technologies in satellite
control systems are low cost and less time consumable design
and build cycle. The onboard autonomy is one of approaches
to decrease the cost of new programs specially in satellite
constellation. An autonomous space system has important
features such as ignoring time consumable activities that must
be done by human [1], cost reduction in operational levels,
reliability enhancement by continuous system health
monitoring, operational optimization by faster reaction and
interrupt ignorance, system protection against system’s
internal issues, compatible and reliable functionality in
complex tasks. Autonomy in satellite is defined as a set of
onboard activities and capabilities that is used to manage
satellite and its payload automatically. Autonomy in satellite
can be done by prolonged onboard planning and scheduling,
remote command generation, onboard attitude determination
and control, independent payload function, event-based
functioning, separation in TT&C and mission control centers,
telemetry data compression, goal based and onboard mission
planning and scheduling, autonomous mission generation and
autonomous fault detection, isolation and system recovery [2].
These mentioned features enable the onboard systems to
decide precisely while confronting unexpected events. This
type of system has automatic reaction capability to reduce or
ignore effects of unexpected events in mission completion.
Usually satellite controlling is done from mission control
center. One of the tasks of this center is to generate a time
sequence of satellite functionality. This time sequence is
called timeline. One timeline can support days or hours of
satellite functions. The obvious problem of this structure is the
limitation on environmental changes. These changes can be
internal like hardware malfunction of payload or be external
like a goal point being camouflaged by a cloud during imaging
task. To conquer these kinds of problems, the onboard
autonomy structure uses a planner-scheduler. Since in
autonomous system some routine tasks of non-autonomous
systems are omitted, utilizing this system can reduce the
complexity of mission control center. To monitor and track
system functionality there is no need of a human operator, this
operator just acts in unexpected or critical situations. For
instance, the accurate analysis of telemetry data is done by a
human operator. One of the other advantages of autonomous
systems is the high ratio of useful data to all volume of data
sent from or to the satellite. In non-automatic methods, most
part of data that are sent to the satellite is software control
statements while most part of data in autonomous methods is
just mission data.
Considering above mentioned advantages of using
autonomy in satellites, designing autonomous systems are a
high priority subjects of every space research centers [3]. The
greatest challenge in design of this type of systems is
developing a heuristic architecture that can be capable of
conquering mission requirements simpler and more effective.
These architectures are exactly against ground based
architectures which are high cost and also having no
permanent monitoring system can cause major failures in their
lifetime. In this article, first an analytic discussion about
178 978-1-4244-9615-0/11/$26.00 ©2011 IEEE