Hello
I am an engineering student and for my final year project I am looking to develop a solar powered drone. The aim is for the drone to charge a small battery/capacitor using an attached solar panel whilst on the ground. Once it detects it has enough charge it will fly to the pre-set location, land and charge up again for the next flight.
I am thinking of using an APM 2.6 or Pixhawk for the project..
How feasible would this be?
Replies
Better yet, I have found this epoxy panel:
http://www.ebay.co.uk/itm/10W-epoxy-Solar-panel-10-Watt-12-V-with-D...
It is 10W and weighs 363g. I could potentially put two in parallel which could mean an output current close to 1A.
We discussed that here...
http://www.rcgroups.com/forums/showthread.php?t=2292431
Solar is a bust for most applications, but in this contrived scenario it could work just fine.
Let's say you carried this 22" x 14.5" thin film panel here...
R7 7 Watt Rollable Solar Panel
It's 269g, so that's reasonable. Maybe you could roll and unroll it somehow, or dangle it beneath. You could also probably do a standard mount if you had 14.5" of clearance between your rotors.
It puts out 7W. At a 12v charge voltage that's .58A (580mA).
So with a 2200 mAh battery pack it might charge in around 4 hours (3 hrs. 48 min, if optimal) if you had direct, unclouded sun.
Depending on your disarmed power draw, if you had good sun, you might be able to charge without even turning your gear off.
In any case it would be pretty trivial to design a startup circuit for your device. The solar panel can stay connected all the time. All you need is a cutoff circuit that disconnects the main power and turns back on after a pre-programmed time or after reaching a certain voltage on the LiPo.
If you can get 10-15 min flight time then you could fly one-way for 15 min, charge 4 hours, fly another 15 min, charge, fly, etc..
So in a 12+ hour day (summer time) you could theoretically get 4 flights a day with a system like I just described.
In northern Alaska the sun is up for 85 days straight from May 10 to August 2. In that situation you have:
85 days x 24 hours = 2040 hrs.
2040 hrs. / 4 hrs. charge = 510 flights.
On either end of that you probably have a lot of very long days also!
Anybody want to run the numbers with real world flight times and ranges? How far could you get in 510 flights? I see the potential to set some crazy records!
I bet the solar company would be happy to sponsor anyone with the balls to try and pull something like this off!
That rollable solar panel looks interesting! Not something I originally considered. Finding a way to roll and unroll it may prove difficult however.
The approach I was thinking of taking was on a smaller scale. A frame similar to this:
http://www.hobbyking.co.uk/hobbyking/store/__64615__HobbyKing_FPV25...
This would allow me to put the electronics inside the frame and mount the solar cells on top.
For solar cells there seems to be an abundance of small, cheap cells available such as:
http://www.dx.com/p/lx-b580-0-8w-pet-laminated-solar-monocrystallin...
I could connect several in series/parallel to obtain a suitable power output. Alternatively, I could mount a large panel which will cover the whole quadcopter as well as the props. I am unsure just how much this will affect the flight characteristics.
You think "dangling solar cells under below the quad" will not have a big aero dynamic penalty??? Sure, other than acting like a giant overweight sail making it completely unflyable.
I don't think you'll get much usable power out of the surface area of the frame or the area available on top.
Your best bet is dangling the solar cell below the quad. Then you can get unlimited area with no big aerodynamic penalty. Putting anything above or below your props is going to completely make your craft unflyable.
Good points above about the solar angle, etc.. But the base numbers show it should be doable. If you have to deal with Z,Y, and Z penalty then oh well.
I'd think that if you could get even 1 flight per day it would be an amazing achievement. There's always larger solar cells and any number of possibilities for pointing them towards the sun.
well, with 12h days, you probably could make 3 flights max, with a full battery on the start of the day, and 2 full recharges. you never get the max power output of the cells, and on the start and end of the days, unless you have a method to track the sun with your panels, the angle will get you almost nothing of energy.
the same goes to alaska, 85 days of sun, but its so low on the horizon that you´ll get trouble to get that energy. weather in alaska is also known to not be the best to carry a big and fragile panel around.
if its for the concept, go ahead. but maybe you can consider a blimp instead of a quadcopter, it already has a big surface area, the power consuption is also very low, and you can even rotate the blimp to achieve a good position for the panels.
The amount of power these devices use is not in line with the amount of power tiny little solar panels would output to charge the batteries. It would take eons to charge a 4000mah battery using the little solar panels small enough to actually attach to it.
I was thinking of having a very short flight time. Initially it may only 'hop' a few meters in one direction. This will mean a considerably smaller battery can be used (maybe even a capacitor?), saving on weight.
I will need to have a look into the solar cells available to see if this is possible.
Nice idea. The way you have described this, it may just be on the outer edge of feasability. Are you thinking fixed or rotary wing? I think your biggest problem with fixed wing will be the autonomous landing/launch cycle (this may well prove to be a project killer). With rotary, you would probably have to devise a way of stowing the solar cells for flight.
I can see this working, but it will not be easy at all. I'd see this as being more suitable as a project for a university engineering team, rather than one student. Best of luck.
The idea is to use a quadcopter. I agree the solar cells will have to be arranged in such a way to give a large surface area whilst not affecting flight performance considerably.
I have not used ardupilot before but it looks to be the best platform out there at the moment for use in this project.
-
1
-
2
-
3
of 3 Next