Environmental Engineering Reference
In-Depth Information
a good point in the timeline to insert the dump, and would order execution
of the dump by the ACS at the appropriate time.
Further, in the event of the detection of onboard anomalies, the Adaptive
Scheduler could take corrective action, which might involve deleting a target
from the list, temporarily skipping a target until it could be observed at a later
date, or even more elaborate reordering of the ground-supplied target list. For
example, if after the ACS had exhausted all its acquisition options the JWST's
fine guidance sensor had still failed to acquire a guide star, the Adaptive
Scheduler could command a deletion of that target from the observation list
without waste of additional observing time and could order that a slew to the
next target on the list be initiated.
For JWST's L2 orbital geometry, an event-driven scheduling system would
be simple, ecient, and easily responsive to at least a substantial list of pos-
sible anomalies. A LEO orbit would present much more of a challenge due to
the complexity of the environment and would probably require the support of
an elaborate look-ahead capability.
In one respect, adaptive scheduling is more complex than the absolute-
timed paradigm. Because under the old paradigm, the start and end of on-
board activities were very well-defined, the process for insertion of realtime
commands into the timeline (provided ample uplink opportunities were avail-
able) was well-defined, if at times awkward. However, under the new approach,
ground operations staff will not always be certain when “safe” opportunities
for realtime uplink might present themselves. Therefore, additional “intelli-
gence” would need to be present onboard to enable the FSW to consolidate
information and prioritize commands from a wide variety of realtime and
preplanned/predicted sources, such as uplinked realtime commands, realtime
sensor output, realtime event messages from onboard performance monitoring,
and the ground-supplied target list.
The FSW must still support ground uplinks of time-tagged commands
along with its list of event-driven activities for those situations where an
activity must be performed within a specific time window (for example, a
ground-planned orbit stationkeeping maneuver). This need is addressable via
a short-term look-ahead capability that would support delaying scheduling
any timeline events that would “swamp” the absolute-time window until after
the absolute-timed event has completed, and giving the absolute-timed event
priority over any routine engineering event that needs to be inserted.
It is apparent just from the Adaptive Scheduler example that consider-
able gains in eciency can be realized just by moving from a fully ground-
programmed, absolute time-driven style of operation to a ground-ordered,
onboard event-driven style. Exploiting the onboard system's knowledge of re-
altime information enables a flexible response to on-orbit events that could
greatly reduce loss of valuable observation time arising from conservative
time-padding in ground look-ahead models. Nonreplaceable onboard resources
could be utilized more economically by expending them in response to re-
altime measured needs, while renewable resources could be allocated more
Search WWH ::




Custom Search