Hi Jun,
Thank you for your patience as we’ve reviewed the latest FreeFlyer files you sent us. The particular engineer who was reviewing your questions is out of the office sick today, so it will be another day or two before our full analysis can be sent back to you. However, he conveyed the preliminary results to us yesterday which I will share with you now:The primary engineer who’s reviewing your questions is out of the office sick today, so it will be another day or two before our full analysis can be sent back to you. However, they conveyed the preliminary results of their analysis to us yesterday; which I have bulleted below:
• The primary reason you are seeing the discrepancies is simply due to lack of tuning in the Kalman filter. In the current configuration, much of the tracking data is being rejected, initially, during the Mission Plans set up. Fortunately, this is a simple fix and if changed you will see better results.
• Your Batch Least Squares process is currently set up as follows:
While(BatchLeastSquaresOD1.IterationCount < 0.5) then; Break; End;
End;
Though this does work, the preferred (and simpler) method is to use the following, which should also result in …show more content…
Even the smallest of differences in initial states will grow over time when numerically integrated. From our experience in operating LEO satellites for NASA, we use much shorter state propagation times, which helps OD results stay more accurate. For example, it is common for missions to compute 7-day ephemeris files on a daily basis to ensure overlapping data in the ephemeris throughout the mission. Then, once a month, compute a 4-6 week ephemeris for science planning purposes. This helps keep your OD data accurate and mitigates slight errors that would acrew over long periods of time and dilute your
Thank you for your patience as we’ve reviewed the latest FreeFlyer files you sent us. The particular engineer who was reviewing your questions is out of the office sick today, so it will be another day or two before our full analysis can be sent back to you. However, he conveyed the preliminary results to us yesterday which I will share with you now:The primary engineer who’s reviewing your questions is out of the office sick today, so it will be another day or two before our full analysis can be sent back to you. However, they conveyed the preliminary results of their analysis to us yesterday; which I have bulleted below:
• The primary reason you are seeing the discrepancies is simply due to lack of tuning in the Kalman filter. In the current configuration, much of the tracking data is being rejected, initially, during the Mission Plans set up. Fortunately, this is a simple fix and if changed you will see better results.
• Your Batch Least Squares process is currently set up as follows:
While(BatchLeastSquaresOD1.IterationCount < 0.5) then; Break; End;
End;
Though this does work, the preferred (and simpler) method is to use the following, which should also result in …show more content…
Even the smallest of differences in initial states will grow over time when numerically integrated. From our experience in operating LEO satellites for NASA, we use much shorter state propagation times, which helps OD results stay more accurate. For example, it is common for missions to compute 7-day ephemeris files on a daily basis to ensure overlapping data in the ephemeris throughout the mission. Then, once a month, compute a 4-6 week ephemeris for science planning purposes. This helps keep your OD data accurate and mitigates slight errors that would acrew over long periods of time and dilute your