I'm developing a Kalman filter-based GPS/INS integration system, and I'm worried about one of my core assumptions in testing. I've been assuming that the u-blox 5 GPS produces a 4 Hz true fix (v5.0 firmware), with one-sigma RMS position error of 10 m and 0.1 m/s of velocity error (based on true dopper rates). Are these assumptions reasonable based on the actual observed performance of the modules?
Thanks, -Jonathan
You need to be a member of diydrones to add comments!
The UBlox outputs a position accuracy estimate in the binary protocol. I am planning on using that for the position error sigma. I was also assuming the same velocity error of 0.1 m/s.
I am only going to use the GPS data at 1 Hz since the inertial drift difference between 250 ms and 1000 ms is negligable. I don't think you will get any better performance out of the GPS/INS running it at a higher rate. My $0.02.
Replies
The UBlox outputs a position accuracy estimate in the binary protocol. I am planning on using that for the position error sigma. I was also assuming the same velocity error of 0.1 m/s.
I am only going to use the GPS data at 1 Hz since the inertial drift difference between 250 ms and 1000 ms is negligable. I don't think you will get any better performance out of the GPS/INS running it at a higher rate. My $0.02.
Tom