#11 Help modeling the gravity model used in EOM
I am working on creating a navigation EKF that relies on data from a strapdown INS and is bounded by position and velocity updates from a GPS. Using the INS measurements (which I take from JSBSim) I would like to predict the position of the aircraft within a 1m radius sphere. I have ran into trouble on what to use for my gravity to update the accelerations in the NED fame. I know JSBSim utilizes the WGS84 equation, but when I use that in my navigation equations the position I calculate doesn't match the JSBSim position outputs. Is JSBSim using geodetic or geocentric latitude to calculate gravity? Is there more than one WGS84 Equation for gravity? Has there been any errors report or updates to the gravity model after JSBSim-1.0? I also can't seem to find where in the source code gravity is calculated.
I've also tried using the spherical gravity model by setting property/gravity-model = 0 but that doesn't seem to change anything. Am I missing something?
As the simulated INS data I am using a-pilot-x-ft_sec2, a-pilot-y-ft_sec2, a-pilot-z-ft_sec2, pi-rad_sec, qi-rad_sec, ri-rad_sec. Is there any property that will allow me log the gravity at that position for each time step?
Has anyone ran into a similar problem? Can you think of a simple solution I'm missing?
Thanks
JSBSim is using geocentric coordinates to compute the J2 gravity model. The computations are made in the file src/models/FGInertial.cpp
You can get the acceleration magnitude from accelerations/gravity-ft_sec2. I would not use the a-pilot-[xyz]-ft_sec2 properties since they are computed in the body frame and are therefore affected by Coriolis and centrifugal accelerations. I suppose that you want to integrate the accelerations in the inertial frame ?
No errors have been reported so far about the gravity model. However if you might be more specific about your findings that might help to solve your issue and fix the bugs if any. Please describe the result you obtain from JSBSim, how you obtained them as well as the result you are expecting.
Also for clarity, please avoid TLAs (Three Letters Acronyms) such as INS and EKF as this will make your statements clearer to everyone.
Thanks.