Servers by jDrones

Soaring/Gliding for Planes

arduplane

(Marco Robustini) #61

@tridge i’ve installed a real airspeed but is in the flow of the propeller, I’ve no alternative to install it in another location in this glider.
So for this reason the wind estimator and the airspeed autocalibration goes crazy after the takeoff in “Manual” mode, and then the SOAR It fails to behave properly because it finds an unreal estimated wind.
So I ask you to include a change to an existing parameter in the code so we don’t have to add another, that when the throttle is at minimum position (motor off) evaluates the wind and calibrates the airspeed, but when the propulsion is active (throttle different to min position) the code doesn’t have to evaluate the wind and don’t calibrate ther airspeed if the autocalibration is enabled, so it goes virtually switch in the synthetic when the propeller work, or better simply don’t use the airspeed.
In this way both wind estimation and airspeed autocalibration may work, as they elaborate data only when the propulsion is off:

Example:

ARSPD_USE,0 (don’t use the airspeed)
ARSPD_USE,1 (use the airspeed)
ARSPD_USE,2 (use the airspeed only if the throttle is at minimum, and with throttle different to min position use always the synthetic or simply disable the aispeed use)

Here some logs of yesterday, but with the airspeed in that position I think it is useless to analyze them, so I expect the code change if possible.
This could bring benefits to all those aircraft where airspeed could be adversely affected by rotating of the propellers.

https://drive.google.com/drive/folders/0B3HEcrNCOga9Sk00aFFFNEFva28

Cheers,
Marco


(Marco Robustini) #62

Really nice result after the special patch by @tridge for my glider with the aiirspeed in the propeller flow (work superfine Andrew, if you want push it on the master!).
Virtually does exactly what I asked in my previous post, with the parameter “ARSPD_USE” to 2 the airspeed work only if the “CH3 IN” is at min value, otherwise, the EKF estimates the wind according to various calculations without using the real airspeed if the propeller is rotating.
Here the tlog and df log of a flight of today with my Elipsoid, bad wheater condition, cold and crosswind, but sometime i’ve found a good sporadic thermal.
Many attempt, in Cruise, FBWB and AUTO mission (3 waypoint in loop).
The only issue is when the SOAR mode auto-switch in Loiter, dunno why it doesn’t circle, takes one direction and keeps it forever.
I’m aware have to look for the thermal, but I think it’s not all in one direction… :joy:
Also when it comes to SOAR_ALT_MIN always runs RTL, i don’t understand when it avoids doing this.
So @Samuel_Tabor please check my logs and if you found any problems please fix it.
My SOAR_POLAR_K is calculated according to my configuration, (16*1.987)/0.49.

https://drive.google.com/drive/folders/0B3HEcrNCOga9Sk00aFFFNEFva28?usp=sharing

Bests, Marco


(Marco Robustini) #63

Today i’ve tuned “well” all the “Polar” parameters for this glider and now i think the SOAR mode is perfect.
I’ve found only one really good thermal condition n this flight and my Elipsoid try to follow it in Loiter, see the new logs on my Google Drive linked in the previous post.
See the tlog at 36%, the Loiter start at 220 meters, at 340 i engage RTL because the glider is out the VLOS… :joy:

Marco


(Samuel Tabor) #64

Hi Marco,

Glad you had some success! I had a look at your log and indeed that thermal was pretty well hooked. I can also see there were quite a few false positives. You can reduce these by improving your airspeed tracking, which in turn relies on the pitch controller. Although the variometer does correct for increases in height at the expense of airspeed it’s not perfect, so variations in airspeed can still cause false positives. This is a feature of the soaring code that could do with some improvement.

Cheers!


(Marco Robustini) #65

Hi @Samuel_Tabor,

I’m doing now some SOAR tests with X-Plane 10 to understand how the code works.
Behavior is completely different from reality, if you look at my Aileron (CH1 and CH5) when enter Loiter they move fast, up and down very quickly and repeatedly, on the SITL instead they act as they should, so there is something that doesn’t work.
Here a screenshot using SITL, there’s clearly a point where it has entered thermal…

Marco


(Marco Robustini) #66

In the first image the AIL channel in SITL during thermal, in the second the AIL channel of my Elipsoid during thermal.


(Marco Robustini) #67

Hi @Samuel_Tabor, no news = good news? :relaxed:


(Samuel Tabor) #68

Hey Marco,

Sorry I haven’t managed to take a look at that issue yet, thanks for the reminder though!

As far as I’m aware there should be no difference between SITL and flight in how the soaring code works. I think the most likely cause is that the higher variance in the real-world measured vertical speed is causing more movement in the loiter waypoint, and the navigation controller is trying to track this, causing more aileron movement. Hopefully I can take a proper look though the log this week and make some suggestions.

Cheers
S.


(wkf94025) #69

I had a whacky Idea for some fun coding: a flight mode for “exit thermal”. Imagine a combination of the following: butterfly to kill lift, rudder in the direction that is most likely to exit the thermal, and down elevator up to the point of Vne, watching max air speed, and holding this configuration until vertical air speed is zero or negative. If the above protocol does not succeed inexiting massive lift, DefCon 1 would be “bury the sticks”, i. e., hard rudder and elevator. What do you think?

Kelly


(wkf94025) #70

It looks like I’ll be able to get the ArduSoar code flying on a couple of my gliders soon. After many months away from RC, I’ve downloaded the latest MP, and ArduPlane, v3.7.1. Is the soaring code in there? Or do I need to use some other beta version or other branch of the code to try the ArduSoar code?

Kelly


(wkf94025) #71

Ignore my post from yesterday. I downloaded the latest Arduplane beta, and the soaring parameters appear now in MP.

Kelly


(gmorph) #72

All the SOARing parameters are now available in the latest 3.8 released version of ArduPlane. You first need to set SOAR_ENABLE to 1, then reboot and then the parameters will be available.

Thanks, Grant.


(wkf94025) #73

Looking forward to trying out the soaring code in 3.8. Unfortunately I am waaaay too deep into aerial mapping with Pix4D, initially with Phantom 4 Pro, but now launching a MyTwinDream build that will run Arduplane 3.8, making use of terrain following for some large mapping missions, and hopefully incorporating Reach RTK GPS. I might be a while in this current bent, but look forward to ArduSoar shortly after that.

Kelly


(Marco Robustini) #74

Hi @Samuel_Tabor, please read this:

https://github.com/ArduPilot/ardupilot/issues/6773


(Adam) #75

Any news in this topic ? I’m realy interested in soaring functions. For this UAV motoglider:

Plane is designed for using a termal conditions so automatic functions for this should do a really very long flights.


(gmorph) #76

ArduPlane supports Soaring. Install an Autopilot and let us know how you go. :slight_smile:


(Gary Mortimer) #77

I would say the update is, yes it works :wink: