- apply sensor scaling immediately to keep things simple (FIFO vs regular)
- inline filter helpers (minor performance improvement)
- dynamic notch filtering
- reorder by axis (applied per axis)
- don't remove notch filters immediately if ESC or FFT data times out
- constrain notch filter frequency and bandwidth to safe range (minimum bandwidth for flaot precision, Nyquist, etc)
- add safe constraint on dt
- merge NotchFilterArray into regular NotchFilter (apply vs applyArray)
- only use direct form 1 to prevent reset confusion
- safe default field initialization
- update VehicleAngularVelocity usage
- use WelfordMean library to determine sensor sample rate average
- if gyro samples exceeds queue depth, instead round to nearest even integer to ease scheduling
- force update if publication latency is falling behind
- remove redundant sample interval perf counters
- status print accel & gyro interval standard deviation
-constrain range to -20..35°C
-do not use temperature readings if exactly 0 (as likley not filled by driver at all in this case)
Signed-off-by: Silvan Fuhrer <silvan@auterion.com>
In airspeed-less mode, instead of the average of min/max airspeed, take the cruise
airspeed (from param) for current airspeed estimate. A diff of the airspeed setpoint
from this value results in increased enegery demand (incerase of throttle) even
in airspeed-less mode.
Signed-off-by: Silvan Fuhrer <silvan@auterion.com>
- this fixes a minor edge case when you only have a single accelerometer and haven't made any parameter changes, the sensor sample rate (used for filtering) might not have been updated from default
- this error is primarily useful when significant gaps in magnetometer are present, otherwise the occasional gaps at startup, etc are distracting in regular usage
This includes the following:
- Remove warning/throw_error flag as it is always !home_alt_valid.
- Remove impossible code path when home_alt_valid = false and
throw_error = false.
- Add home_alt_valid check in second check before using home_alt.
- Only use warning for the second check to allow waypoints below home.
- Don't return early when only warning.
This adds support to handle INT32_MAX for COMMAND_INT.x/y by converting
it to NAN internally.
It also adds paranoid checks to prevent:
- NAN being used sent by a COMMAND_INT by mistake.
- INT32_MAX being used sent by a COMMAND_LONG by mistake.
e.g.:
WARN [mavlink] stream ADSB_VEHICLE not found
WARN [mavlink] stream GIMBAL_DEVICE_ATTITUDE_STATUS not found
WARN [mavlink] stream GIMBAL_MANAGER_STATUS not found
WARN [mavlink] stream GIMBAL_DEVICE_SET_ATTITUDE not found
WARN [mavlink] stream GPS2_RAW not found
WARN [mavlink] stream UTM_GLOBAL_POSITION not found
ERROR [mavlink] configure_streams_to_default() failed
The calibration is not found in HITL mode. Therefore, I suggest to
ignore this step and assume the calibration is fine.
This mostly fixes the preflight check indicator in QGC, arming was (for
some reason?) already possible.
Daniel Agar