Extract attitude data. Swift attitude data appears
in several places in the telemetry. This subroutine
collects all the attitude data into a single quaternion-
based attitude file and removed redundant information.
The main source of attitude data is the "ACS packets"
(APID 404). The BAT also places ACS records in the
headers of each LDP and in the body of a special
LDP generated suring slews.
The XRT places ACS records in the header for each CCD
frame, however these are represented by "floats" instead
of "doubles", so they are not used if higher accuracy
data are available.
The sub checkDatabaseForNominalPointing reads the database
file "/sw/nominal_pnt.rdb" from the directory indicated by
env var $RDB_TABLES_DIR (normally set in aps.config;
defaults to /aps/db/rdb/tables). If the sequence is found
in this table, those values are used for the nominal
pointing; this allows for manual override when the pointing
cannot be reliably determined automatically. Otherwise, the
mean pointing is calculated and compared with the requested
pointing (SACSREQ, read from the engineering housekeeping
file), if these disagree an E2 error is logged and an email
notification is sent. Currently the tolerance is 10 arcmin.
Someone should then review the sequence and add the correct
position to nominal_pnt.rdb, then rerun the sequence.
Here is a pseudocode outline of the algorithm used to
determine the nominal attitude:
IF obs in ADB [nominal pointing database]
pnt_db = ADB[obs]
ELSE
pnt_sacs = determineSACSREQ(sw<obs>sen.hk)
# 1-2% of the time, determineSACSREQ fails...
pnt_aspect = aspect(sw<obs>[pu]at.fits)
IF pnt_sacs AND pnt_aspect THEN
IF separation(pnt_sacs, pnt_aspect) > 10 arcmin THEN
E2 ERROR SACSREQ and aspect positions inconsistent
SEND EMAIL NOTIFICATION
ENDIF
IF minRollDelta(pnt_sacs, pnt_aspect) > 2 deg THEN
E1 ERROR
ENDIF
ELSEIF pnt_aspect
IF pnt_aspect has ERROR
E2 ERROR Unable to determine nominal pointing
SEND EMAIL NOTIFICATION
ENDIF
ENDIF
ENDIF
Parent Classes:
Subs::NominalPointing, Subs::SwiftSub
Constructor
Accessors
Methods