Closed TSC21 closed 10 years ago
@vooon @TSC21 You can always send yaw using the SET_LOCAL_POSITION_SETPOINT message if you want to use local planners. The message is already implemented in the setpoint_position plugin, we just need to add a way of sending the yaw (custom message?)
That plugin is already created, Kabir: it's setpoint_attitude.
---Nuno MarquesAlferes Aluno de TransmissõesAluno de Mestrado em Engenharia Electrotécnica e de ComputadoresInstituto Superior Técnico | Universidade de Lisboa +351 912 090 991Azambuja | Portugal Dreams set what you want. Actions determine what you conquer!This message and any files herewith attached may contain confidential or privileged information and is intended solely for the use of the entity to which it is addressed. If you receive this message in error, please notify the sender immediately and delete this message and any files attached without copying them in any way.
Date: Sun, 27 Jul 2014 10:01:21 -0700 From: notifications@github.com To: mavros@noreply.github.com CC: n.marques21@hotmail.com Subject: Re: [mavros] ATTITUDE_SETPOINT_EXTERNAL msg plugin (#64)
@vooon @TSC21 You can always send yaw using the SET_LOCAL_POSITION_SETPOINT message if you want to use local planners. The message is already implemented in the setpoint_position plugin, we just need to add a way of sending the yaw (custom message?)
— Reply to this email directly or view it on GitHub.
We don't actually need it for local planners... Thats what I'm saying. We could just use the same setpoint_position plugin for that.
We have too many setpoint_* plugins already... I think we just need some ones :
setpoint_position (XYZ position + Yaw) setpoint_orientation (Yaw, Pitch, Roll, Thrust)
Rest are a bit superfluous in my opinion. They do have usefulness in many cases, but I think those use-cases are too user specific. Adding support for every single mavlink setpoint message in mavros and yet having half-baked support on PX4 side makes no sense. We should do this in a structured way. Let the extra plugins be available in a seprate repo for those who will be using them, but adding such specific plugins will not be useful for 80% users.
Kabir
I agree partially. In my case, my controller is implemented using velocity control, and not position. So a setpoint_velocity is needed. Lacks implementation on PX4 side but indeed is going to be usefull to me also. Is structured after all cause it goes to people needs. You cannot say who needs this or that so I think giving options for control setpoints it's a way of guaranteeing flexibility of the bridge.
If you check we are creating plugins to the new msg API, so it's a plugin to LOCAL_NED_POSITION_SETPOINT_EXTERNAL and not to SET_LOCAL_POSITION_SETPOINT, which in this last case send XYZ and yaw as you are saying. It's a mavlink msg dependent plugins we are creating and that's the most structured it can get. And since Yaw it's send in ATTITUDE_SETPOINT_EXTERNAL, then a plugin for that is needed.
The only thing we need now is help on PX4 side or get in an adventure figuring out what to code.
@mhkabir I agree, is too mush setpoint plugins, probably better merge some of that in one. But in one hand i get harder support and in other KISS.
@vooon there was the option I gave you for using a parameter in a unique setpoint point plugin where you choose if the user want to use velocity control, position control or accel/force control. I warn that in PX4 side already position and force control msgs are handled in mc controllers.
--- Mensagem Original ---
De: "Vladimir Ermakov" notifications@github.com Enviado: 28 de Julho de 2014 11:41 Para: "vooon/mavros" mavros@noreply.github.com Cc: "TSC21" n.marques21@hotmail.com Assunto: Re: [mavros] ATTITUDE_SETPOINT_EXTERNAL msg plugin (#64)
@mhkabir I agree, is too mush setpoint plugins, probably better merge some of that in one. But in one hand i get harder support and in other KISS.
Reply to this email directly or view it on GitHub: https://github.com/vooon/mavros/issues/64#issuecomment-50323626
With this we can also control the Yaw of the quad, which is recommended to use in motion planners. Roll and Pitch also, but that depends on the type of controller it's being used (probably using a velocity control it's easier when having a velocity estimation).
I'm going to start a setpoint_attitude.cpp. I'll send you when I have it ready.