Kalman Filter Input Processor for Boresight Calibration
- Created: Wednesday, 01 January 2014
- NASA’s Jet Propulsion Laboratory, Pasadena, California
The new software brings this technology to the industrial level.
Ka-band ranging provides the phase center (PC) to phase center range, which needs to be converted to the center of mass (CM) to center of mass range. Nominally, both PC and CM lie on the line connecting the spacecraft GRAIL A and GRAIL B. In this case, the conversion should be done simply by adding the CM-to-PC distance L to the measured range for both spacecraft. However, due to various technical reasons, such as displacement of the true CM from its nominal position in the SRF, or spacecraft attitude fluctuations, the PC and CM define a unit vector that may be different from the nominal line of sight. The objectives of the software are to determine the actual line of sight direction for each spacecraft and correct the previously recorded range data, and to provide instructions for how to maneuver each spacecraft to make necessary attitude corrections.
While elements of this approach have been used for the boresight calibration in the GRACE project, the new software brings this technology to the industrial level. It is now fully documented and can be used by people other than its developers. This innovation provides graphic outputs and log files that are critical for quick analysis and troubleshooting. In addition to the line of sight direction, the software allows one to evaluate the CM-PC base length, which is important when the PM location is subject to variations (e.g., due to fuel depletion).
This software is implemented in Python and offers excellent cross-platform porting possibilities. It is very versatile, and may be applied under various circumstances and for other related purposes. This innovation is capable of combining the input data from several calibration maneuvers, evaluating individual range biases, and compressing the time stamps. It uses Lagrange interpolation for the orbit data, and a unique quaternion-interpolating algorithm for interpolating the attitude data. As a result, data files with different data rates and independent time stamps can be handled together.