Add influence of wind to rotor drag

[Jaeyoung-Lim]

This PR adds the influence of the wind to the motor model, by considering wind as part of the calculation of rotor drag. This is important since the majority of the drag is usually generated from the rotors in case of multirotors, and is influenced by wind. Thrust variations are not considered.

Also, @MaEtUgR have brought up examples where some issues only appear under the influence of wind. This PR would make it easier to see wind influence of multirotor models.

Testing
Below shows the model taking off and being influenced under the wind of 5m/s

Additional Context
This is a follow up PR of #375 , which added wind influence to the lift drag model

[LorenzMeier]

@MaEtUgR Could you please review and approve? We need to get faster in closing things out.

[MaEtUgR]

@LorenzMeier I really try to follow up as quickly as possible but I was out for the last 3 days over the weekend and my day off and looking into it now.

[MaEtUgR]

It flies somehow strange compared to the pure force wind against the base_link. I have to do some more testing tomorrow.

[MaEtUgR]

I did some more testing and it's a matter of magnitude of the disturbance. I think applying the wind velocity like this is correct and the "strange" flying occurs e.g. at 5m/s because the wind is definitely to strong for the actuation of the simulated drone. Especially the yaw authority is very weak and if the drone hits one or multiple motor output thrust limits the results can look quite wonky because there's no more margin to actuate with but it doesn't crash. A wind speed of ~3.5m/s allows for a bit of margin in terms of navigation without getting too far off the navigation goals and 1.5m/s wind is more of a full yaw authority during hover scenario:

The wind simulation definitely helps to test disturbance rejection and most importantly also corner case handling like when running into the limits.

I was testing with make px4_sitl gazebo_if750a and this snippet added to the if750a.sdf:

<plugin name='wind_plugin' filename='libgazebo_wind_plugin.so'>
      <frameId>base_link</frameId>
      <linkName>base_link</linkName>
      <robotNamespace/>
      <xyzOffset>0 0 0</xyzOffset>
      <windDirection>0 1 0</windDirection>
      <windVelocityMean>1.5</windVelocityMean>
      <windGustDirection>0 0 0</windGustDirection>
      <windGustDuration>0</windGustDuration>
      <windGustStart>0</windGustStart>
      <windGustForceMean>0</windGustForceMean>
    </plugin>

It would be really awsome if the Wind linear_velocity parameter that is already exposed in the UI could change the wind dynamically because then the sdf would not need to be edited and it would allow testing step disturbance reactions in certain scenarios mid flight.

[MaEtUgR]

Thank you @Jaeyoung-Lim! That really helps to find corner cases 👍

[Jaeyoung-Lim]

Thanks, @MaEtUgR !