I am trying to do a OD with BatchLSEstimation for a geo satellite with simulated TDOA data. ( A single set of TDOA data, data captured multiple times with in two days). The estimated orbit has a significantly high inclination compared to TLE orbit. The out of plane error is high too. (eg. The inclination in TLE orbit in eci frame = 0.09 and estimated orbit in the eci frame is 0.6 , values are in radians)
I added force models related to sun, moon, solar radiation, gravity field of order 8. The satellite weight and area may not be exact.
I would appreciate some insight in to reasons for this.
First, are the two orbits used for comparison expressed in the same orbital frame?
I ask the question because TLE uses its own frame which is TEME (True Equator Mean Equinox). Maybe you are just comparing two orbits in two different frames.
Then, do you extract the inclination directly from the TLE or by using the SDP4 model (i.e. using
TLEPropagator class of Orekit)?
Because attempting to directly use TLE orbital parameters without any reference to the
TLEPropagator is prone to errors.
Thank you for replying.
I am using SGP4 model (TLE propagator class). Also both the frames are gcrf frame.
(I am using the Orbit determination example code for the python wrapper.)
tle_orbit_TEME = tle_initial_state.getOrbit()
tle_pv_ECI = tle_orbit_TEME.getPVCoordinates(gcrf)
I do not think the error is due to frames.
I understand a single TDOA measurement set could result in a less accurate orbit. But I would like to know whether there are other parameters that can be changed to improve OD.
0.6 radians error in inclination is a huge error.
Are the stations involved in your TDOA data widely separated or close to each other? Are they close to equator ? How many measurements do you have over the two days span?
Thank you for the reply. I tried with three sites and the error reduced significantly. (The inclination now around 0.002 radians).