that prevents the vehicle from crashing with invalid setpoints or
states.
This broke with #16869 when the scheduling of the position control module
and the setpoint generation got independent. The failsafe mechanism assumed
the setpoint is overwirtten by the possibly infeasible input on every loop
iteration which is not the case anymore. As a result the failsafe reset its
histeresis based on the failsafe setpoint from the last loop iteration.
Keeping the failsafe_setpoint separate solves this issue. Note that
these setpoints to the bare minimum to keep the vehicle safely in the air
and do not suffer from sideffects ignoring the EKF reset.
- handle SET_POSITION_TARGET_LOCAL_NED and SET_POSITION_TARGET_GLOBAL_INT with ORB_ID(trajectory_setpoint)
- FlightTaskOffboard not needed at all
- bypass position_setpoint_triplet entirely (start removing extraneous fields)
- simplify offboard_control_mode to map to supported control modes
In some cases e.g.: (VTOL in wind) a good tracking cannot be
achieved. This condition then needs to be relaxed, otherwise the
drone cannot land properly.
- parameter updates can be quite expensive because they trigger nearly all modules to reload all of their parameters immediately
- limit modules from updating faster than once per second
- split out switches from manual_control_setpoint into new message manual_control_switches
- manual_control_switches published at minimal rate (~ 1 Hz) or immediately on change
- simple switch debounce in rc_update (2 consecutive identical decodes required)
- manual_control_switches logged at full rate rather than sampled at (5-10% of messages logged)
- manual_control_setpoint publish at minimal rate unless changing
- commander handle landing gear switch for manual modes
- processing of mode_slot and mode_switch is now split so we only do one or the other (not both)
- a future step will be to finally drop mode_switch and accompanying switches entirely
Co-authored-by: Matthias Grob <maetugr@gmail.com>
- track and publish validity based on hover thrust variance, innovation test ratio, and hysteresis
- only publish on actual updates or becoming inactive
- fix dt (previous timestamp wasn't being saved)
- use local position timestamp (corresponding) to accel data rather than current time to avoid unnecessary timing jitter
- check local position validity before using
- mc_hover_thrust_estimator: move from wq:lp_default -> wq:nav_and_controllers to ensure the hover thrust estimator runs after the position controller and uses the same vehicle_local_position data
- land_detector: check hover thrust estimate validity and adjust low throttle thresholds if hover thrust is available
- mc_pos_control: only use hover thrust estimate if valid
Before #14212 the velocity control gains used in the multicopter
position controller were defined as a scale between velocity error in
one axis (or it's integral and derivative respectively) and the unit
thrust vector. The problem with this is that the normalization of the
unit thrust vector changes per vehicle or even vehicle configuration
as 0 and 100% thrust get a different physical response. That's why
the gains are now defined as scale between velocity error
(integral/derivative) and the output acceleration in m/s².
- Avoid constantly adjusting the velocity gains with the HTE
- Make sure the hover thrust integral update doesn't break
even though its unit is acceleration and not unit thrust anymore
We need to convert the velocity gains to not contain/depend on the
hover thrust. In horizontal direction it doesn't make sense to scale
them with the hover thrust and in vertical direction the adjustments are
already done in the acceleration to collective thrust conversion.
- create common uORB::PublicationBase
- uORB::PublicationQueued types are now type aliases
- ORB_PRIO use enum type everywhere to avoid accidental misuse
- PX4Accelerometer/PX4Gyroscope/etc driver libs explicitly advertise on construction, unadvertise on destruction. This is a workaround for any potential issues that might appear when accel/gyro cdev and uORB indexing doesn't align.
Because the takeoff ramp is a vertical velocity limit ramp for the
nice user experience but the acceleration feed-forward can
add on top of the output and depending on trajectory generation
result in unwanted thrust changes during the ramp.
Matthias Grob