View Full Version : Discussion Micro Air verhicle Stabilized flying
Maverick2k
Sep 17, 2007, 04:28 AM
Hi,
First of all I want to complement this forum for the lots info you can find.
My question is sort of brainstorm session. I'm working on a graduate assignment for a MAV with a fly mechanism as a fly/butterfly (indoor). My task is to research the possibilities for sensors.
First there will be an investigation about sensors for stabilized flying. Because it is a MAV all sensors must be low power and low weight. So the less sensors the better.
First I want to know the minimum variables I need for a stabilized flight.
The list so far:
Altitude
Leveling
Airspeed
Is there need for more variables?
Youre help would appreciated.
Maverick2k
dmgoedde
Sep 17, 2007, 10:24 AM
for altitude - I know from experience that SCP1000-D01 (sparkfun.com has breakout board for about $50) will fit the bill. sensor itself is about 80mg mass, very low power, 3V device, and I've been able to use it in real-world result of controlling altitude to +/-1 meter, I would use it indoors. I've locked altitude on a plane at head-height, then over 5 minutes plane never got below 2' AGL or above 8' AGL. I can send you some code for Parallax Propeller that does all this.
I don't know much about pitot tubes, but think they are barometer based as well... maybe you can make one from another SCP1000?
for levelling, I know thermophile sensors can be very light, but probably wouldn't work indoors since there is no horizon/sky line.
If your plane is dynamically stable, then you solve the levelling problem for roll, and pitch will take care of itself with proper balance... using SCP1000 for altitude control then you have a complete solution for your 3 requests.
Maverick2k
Sep 18, 2007, 03:18 AM
for altitude - I know from experience that SCP1000-D01 (sparkfun.com has breakout board for about $50) will fit the bill. sensor itself is about 80mg mass, very low power, 3V device, and I've been able to use it in real-world result of controlling altitude to +/-1 meter, I would use it indoors. I've locked altitude on a plane at head-height, then over 5 minutes plane never got below 2' AGL or above 8' AGL. I can send you some code for Parallax Propeller that does all this.
I don't know much about pitot tubes, but think they are barometer based as well... maybe you can make one from another SCP1000?
for levelling, I know thermophile sensors can be very light, but probably wouldn't work indoors since there is no horizon/sky line.
If your plane is dynamically stable, then you solve the levelling problem for roll, and pitch will take care of itself with proper balance... using SCP1000 for altitude control then you have a complete solution for your 3 requests.
Thanks for the information.
The SCP1000-D01 is indeed a suitable candidate. All code is welcome.
rblilja
Sep 18, 2007, 03:51 AM
Have you considering the usage of ultra sound ranging for altitude measurement? Probably the power consumption and the weight will be the problem with that solution. Currently I’m working on a project involving a small helicopter for indoor autonomous flight, my intension is to test the idea.
dmgoedde
Sep 18, 2007, 01:11 PM
Have you considering the usage of ultra sound ranging for altitude measurement? Probably the power consumption and the weight will be the problem with that solution. Currently I’m working on a project involving a small helicopter for indoor autonomous flight, my intension is to test the idea.
Maybe the "Ping" sensor from Parallax.com? I haven't tried it, but it came to mind.
_helitron_
Sep 18, 2007, 01:25 PM
I use such sensors (exactly the SRF10 from here: http://www.robot-electronics.co.uk/shop/Ultrasonic_Rangers1999.htm) since several years for my little robots and in my opinion sonar is not convenient for such an application. The problem is simply the reflections from all the surfaces around. Much better are the IR distance sensors from Sharp but I don't know if the measurable distance of the sensors is big enough for this application. Here for example a site with infos about the Sharp sensors:
http://www.acroname.com/robotics/info/articles/sharp/sharp.html
and here all sensors on a site
http://www.acroname.com/robotics/parts/c_Sensors.html
//Erwin
Maverick2k
Sep 18, 2007, 03:30 PM
Thnx for the ideas,
I will note all the mentioned sensors. And explore there possibilities.
Does anyone know any readable documents over designing an autonomic control for indoor MAVs? I have found some documents with heavy mathematics.
_helitron_
Sep 18, 2007, 06:03 PM
I fear it IS very difficult stuff :D !
By the way Maverick2k, has your planned task anything to do with the indoor flight session of MAV07 ?
http://www.mav07.org/competition_info.html
//Erwin
zik
Sep 18, 2007, 08:08 PM
Much better are the IR distance sensors from Sharp but I don't know if the measurable distance of the sensors is big enough for this application.
Hi Erwin,
I've used the Sharp IR distance sensors and their range maxes out at a meter or two. I think this would probably be too short for Maverick2k. There are commercial autopilots out there which use ultrasonics for landing sensors but I haven't tried it myself.
Cheers,
Zik
_helitron_
Sep 19, 2007, 01:47 AM
Hi Erwin,
I've used the Sharp IR distance sensors and their range maxes out at a meter or two. I think this would probably be too short for Maverick2k. There are commercial autopilots out there which use ultrasonics for landing sensors but I haven't tried it myself.
Cheers,
Zik
Hi Zik,
so far I remember there is a Sharp IR-sensor type with 1.5 or 1.8 m (5 - 6').
Really, commercial autopilots with US-sensors, do you mean fullsize or R/C ? So far I know are US-sensors absolutely useless over grass for example.
I thought more for a little laser distance measuring device, in the meantime there are very little relatively inexpensive devices on the market (from Bosch for example) but I don't know, if this thingies have an interface to the outside world.
Cheers,
//Erwin
zik
Sep 19, 2007, 02:09 AM
Really, commercial autopilots with US-sensors, do you mean fullsize or R/C ? So far I know are US-sensors absolutely useless over grass for example.
I was thinking of Micropilot's AGL. This is what their web site says:
The AGL board is an optional ultrasonic altimeter that provides high resolution altitude information up to an altitude of 16 feet. The AGL board allows for autonomous takeoff and landing of your UAV.
* Highly accurate above ground up to 15 feet +/- 1”
* Allows for autonomous takeoff and landing
* Allows flare
_helitron_
Sep 19, 2007, 03:48 AM
I was thinking of Micropilot's AGL. This is what their web site says:
I'm impressed (if it's really works :) ) !
//Erwin
Maverick2k
Sep 19, 2007, 03:58 AM
Hi Erwin,
I've used the Sharp IR distance sensors and their range maxes out at a meter or two. I think this would probably be too short for Maverick2k. There are commercial autopilots out there which use ultrasonics for landing sensors but I haven't tried it myself.
Cheers,
Zik
The maximum of two meters is too short, I will check some ultrasonic sensors, these are also relative Low Power.
My task is not directly linked to MAV07. I'm only doing some research on sensors and their possibilities for a MAV
_helitron_
Sep 19, 2007, 04:16 AM
Aha, ok, thanks Maverick2k :) !
//Erwin
_helitron_
Sep 19, 2007, 09:52 AM
Hey Maverick2k,
just found a brandnew Sharp IR distance sensor for a 100 cm - 550 cm range:
http://www.futurlec.com/Distance_Sensors.shtml
//Erwin
Maverick2k
Sep 19, 2007, 10:25 AM
Hey Maverick2k,
just found a brandnew Sharp IR distance sensor for a 100 cm - 550 cm range:
http://www.futurlec.com/Distance_Sensors.shtml
//Erwin
Thnx
Here is a url for information about pitch yaw and roll, some basic stuff
http://tom.pycke.be/mav/70/gyroscope-to-roll-pitch-and-yaw
aghaei
Sep 22, 2007, 03:38 PM
Hi dear
I am a student of field mechanic; and my project is VTOL or VTUAV.
I want to know about the maps of design this VTOL whether the model is new or old.
I am very kind of you
vBulletin® Copyright ©2000-2009, Jelsoft Enterprises Ltd.