You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
While flying a quadplane one v1.9 I noticed that there is always a 1 second gap in the vehicle_rates_setpoint topic at the beginning of the transition. Normally the fixed wing attitude controller should be publishing values. The rate controllers still seem to run.
In the figure above you can see the gap in the rate setpoint topic (top right).
At the beginning of the transition we switch from multicopter attitude controller to fixed wing attitude controller to produce the rate setpoint. However, i don't see the reason yet why it takes the fixed wing attitude controller 1 second to publish to the rates setpoint topic.
I'm not sure, but I guess I'm not surprised given the complexity of the independent logic in each separate controller for coordinating which portion is running wrt vtol state and type.
I'm sure there's a whack-a-mole solution to the bug here if someone wants to temporarily get their head around all the current code and timing long enough to see the problem, but how about something architecturally simpler?
Currently we have fw_att_control and mc_att_control both containing attitude + rate controllers. What if we split these both into separate modules, and then turned vehicle_control_mode into an MC/FW virtual topic that the VTOL code could manipulate independently to control the transition behavior?
While flying a quadplane one v1.9 I noticed that there is always a 1 second gap in the vehicle_rates_setpoint topic at the beginning of the transition. Normally the fixed wing attitude controller should be publishing values. The rate controllers still seem to run.
In the figure above you can see the gap in the rate setpoint topic (top right).
At the beginning of the transition we switch from multicopter attitude controller to fixed wing attitude controller to produce the rate setpoint. However, i don't see the reason yet why it takes the fixed wing attitude controller 1 second to publish to the rates setpoint topic.
@dagar Any ideas?
Log: https://logs.px4.io/plot_app?log=5bca4650-96a8-4f92-8619-b9b1cadccf4a
The text was updated successfully, but these errors were encountered: