The description of ARSPD_USE tells that it is used for enabling the use of airspeed for throttle control. For ICE-powered aircraft, it may be not the best option since airspeed is pretty noisy, and fast throttle oscillations are not good for fuel efficiency and engine stability. So there are many scenarios when ARSPD_USE=0 is an option for normal flight conditions. And I found that when turned off it leads to the inability of EKF3 to use airspeed as a velocity aiding and it falls back to DCM if no GNSS available, but if ARSPD_USE=1 it works as velocity aiding in the same situation. Are there any real reasons for it? Is it a bug or intentional behavior? Can we have separate switches for using airspeed for throttle and navigation?
AFAIK any control is done on inputs processes by EKF so airspeed sensor noise should be reasonably filtered, unless you have incorrectly set EKF parameters.
The problem is definitely not with the noisy airspeed. It is with the ARSPD_USE=0 disabling velocity aiding to EKF. But the solution not to use it is not a good one, since there may be various reasons not to use airspeed for throttle control. So I suggest that it should be two separate switches - one for using airspeed for throttle control (as it is written in description) and the second for using airspeed for EKF velocity aiding.
EKF uses airspeed if it is available. ARSPD_USE switches between command throttle and command airspeed for some modes.
That’s not true. You can easily reproduce behavior using SITL:
- Start the mission with ARSPD_USE=1
- Turn off EKF3 horizontal position and velocity aiding by setting EK3_SRC1_POSXY=0 and EK3_SRC1_VELXY=0. EKF3 will resume aiding IMU with the airspeed
- Switch airspeed use with ARSPD_USE=0
- You’ll get “AHRS: DCM active”, “EKF3 IMU0 stopped aiding” and “EKF3 IMU1 stopped aiding”.
Greetings you as a supporter of the internal combustion engine!
Check out my post here Testing of the aircraft with the HOLYBRO DURANDAL flight controller for resistance to large vibrations . The problem with using an internal combustion engine is its vibrations. I had no problem with the airspeed sensor. I performed a test of the aircraft without measures to eliminate the influence of vibration, my goal is to establish the stability of the control system to the appearance of vibration, it was to check the flight controller for how much it itself is provided to this effect.