Closed bresch closed 3 years ago
How do we know if 0.01 is an appropriate minimum for the bank of 5 EKF yaw estimators. I would like to see some testing with RTK GPS or simulated RTK GPS accuracies using SITL.
I tried a few times in STIL jMAVSim and did not find any issue. The bank of yaw estimators converges properly and the variances don't seem to have any numerical issue. example log: https://logs.px4.io/plot_app?log=78810a15-04ba-4645-882e-2669a15ce968
The yaw estimator should use the same maximum GNSS speed accuracy controlled by the user-defined parameter.