Intuition about GPS Disable

ArduPlane Plane 4.5
1

Hi, I need intuition regarding why UAV in Mission Planner follows exact path of waypoints after disabling GPS but deviates in Flight Gear which is the actual Flight Dynamics Modelling result.

  1. When GPS gets disabled, my intuition is that, the EKF should starts estimating wrong location and the error must goes towards increasing. And this has to be seen in Mission Planner even the Controllers are there. BUT WHEN I PLOT THE POS OR XKF1 LAT OR LNG, THE RESULTS ARE EXACTLY THE SAME AS WITHOUT DOING GPS DISABLING.

  2. I have seen the FDM Model and Controllers part in which I didn’t find anything that can make the path deviate in Flight Gear. The inputs are from the EKF which the main culprit. I did Chatgpt and it says the EKF has a yaw drifting and therefore, we can see the actual path being followed in MP.

Kindly, give me an intuition why this is happeing and what is actual working in EKF and where is it in like the function name & file name so that I can track it.

2

IIRC the default configuration for MP SITL is to use sim state directly.

3

Hi, the Flight Gear uses sim state directly but the input in FDM is from servos that are being generated by Controllers. If the EKF estimation is correct then the error is very less and Controllers will generate less servo value that are being fed to FDM. So, that is mean, the Flight Gear Path doesn’t deviate.

But it is the other wise. The path gets deviated in FG but not in MP.