CRISP Tracking Tests

One of the top science objectives for CONTOUR was the acquisition of high
resolution images of the nucleus of each comet we intended to fly by. 
To achieve extremely high resolution images, the plan was to fly very 
close to each nucleus. However, with the high flyby velocities for these 
encounters, the only way to successfully acquire images at closest approach 
was through the use of closed loop tracking. Closed loop tracking was to 
be accomplished with the CRISP instrument. 

In very general terms, the instrument had the ability to analyze images 
taken only minutes prior to closest approach to determine a revised 
(better than ground-based knowledge) location of the comet. This would 
drive a tracking mirror to keep the comet in the field of view regardless
of the actual miss distance.  It would also drive a closed loop feedback 
with the guidance and control subsystem to control the roll of the 
spacecraft to keep the comet withing the mirror plane. This would cover 
the possibility that Encke's position was out of the nominal predicted 
trajectory plane.  It would also allow CRISP to select one sequence from a 
set of pre-loaded encounter sequences to execute at closest approach.
These sequences were each tailored-designed for different encounter miss 
distance. Collectively, these sequences covered the 3 sigma delivery 
uncertainties from the final trajectory correction maneuver. Analysis of 
the images just prior to closest approach would allow CRISP to select the 
proper closest approach imaging sequence and also know the proper time to 
trigger the sequence.

The encounter sequence could be run either in the open loop or closed 
loop mode. Closed loop mode (described above) could only truly be tested
in the lab (instrument on a pivoting table with a bright target source to
simulate the comet) or at a cometary encounter.  With the instrument on 
the spacecraft we could only test the encounter sequence in open loop mode. 
In this mode, the spacecraft does not analyze images taken on approach. 
Instead it tracks a pre-loaded off-nominal trajectory and makes a selection 
of the proper closest approach sequence, and chooses the proper kick off 
time for that sequence based on that off-nominal trajectory.  In this 
manner we were able to test many of the hardware and software components of 
the process even though we could not fully test the closed loop process. 

The tracking test simulations performed on April 6 and May 16, 2002 were 
series of tests in which we simulated an open loop encounter flyby using 
these off-nominal trajectories as described above.  We would load an 
off-nominal trajectory, initiate the encounter sequence, and wait to see 
which closest approach macro it would kick off, and at what time it kicked 
the macro off. 

As a result of the March 13 MSIM2 Encounter test, a bug was discovered in 
the ground software that converts the Chebychev polynominal trajectories 
into the format used by the spacecraft. We continued to trouble shoot
this problem in addition to running the tests. 

The directory chebys_april6/ contains the actual chebychev files for the
off-nominal trajectorires used for the April 6 tests. The directory 
chebys_may16 contains the trajectories used for the May 16 tests. 

The other files in this directory are the actual sequences that we
used for these tests. Tracktest_short_seqout.asc and Tracktest_long_seqout.asc
were two versions of the encounter sequence ( a long test and a short test). 
We had to create a long test for those trajectories with very large miss
distances because they were expected to kick off at a greater distance
from closest approach (earlier in time). The majority of the tests run
were the short version (for expediency). They are identical except the
long test sets up and initiates earlier relative to closest approach.
 The file tracktest_macros_seqout.asc is a compilation of the different 
closest approach macros we loaded into CRISP for these tests. These were 
the imaging sequences that the TPU had to choose from at closest approach. 

Despite some frustration with getting things to run, the tests were generally
very successful. Except for a number of minor problems that were understood,
the CRISP TPU and mirror tracked exactly as expected, and in each case
kicked off the proper closest approach macro at the proper time.  The
guidance and control also caused the proper spacecraft roll to occur in 
each case.