[Canberrauav] UAV project status
Moss, Andrew Andrew.Moss at ed.act.edu.au
Wed Jan 19 19:11:54 EST 2011
Hi there,
It is good that everyone is back and on-line again (well except for me - NZ). It is
great that the mailing list is up and running as well. There has been a fair bit of traffic
on a number of topics and it would be nice to get some structure to this information
flow. Please feel free to comment on the following suggestions:
1) We should decide on the team leaders for each of the groups. I suggest :
a) Project Management/requirements management/deliverables/schedule - Stephen
b) Airframe/autopilot/failsafe - Jack
c) comms - Dave
d) Sponsors/corporate/promotion - Andrew Moss
e) Video processing/target acquisitions - Tidge (Andrew)
f) Ground Station - ??????
g) website - ????
Remember these are the leaders of the groups. They don't have to be the experts OR do
all of the work. There will be a fair bit of interaction between all of the groups.
2) Do we need separate forums ( mailing lists ) for each group. Is the information in the
Canberra UAV list getting overwhelming already?
3) We should have regular meetings. How about every second Wednesday?
4) Please decide what you want to buy. We are coming to the end of our "technologies
investigations" phase. So far we have bought :
a) 2 x ArduPilots ( integrating currently )
b) 1 pair of 1mW 2.4 GHZ Xbee radios ( working, range is dismal, very limited usefulness)
c) 1 pair of 50 mW 900Mhz Xbee radios ( working, excellent for medium range testing )
d) 1 x 2.4 GHZ 1 watt amplifier ( not tested )
e) 1 x 900MHZ 500mWatt video system with camera and gimbal ( wrong frequency band
for Australia, not useable )
f) 1 x fleet of test airframes ( we have one less now, not sure how many are left )
g) 2 x GPS chips ( Working well, one on autonomous truck, for software development, 1 metre
accuracy, very long startup, connected to google earth, not tried on plane yet)
h) 1 x AVATAR airframe (for display and promotional purposes)
i) 3 x Arduino boards ( three different varieties, all working well, looking at the ECLLIPSE IDE
rather than the Arduino IDE )
5) Website is up and running but really need an web developer.
6) Work to be done over the next week :
a) frequency testing on 900Mhz XBEE. To be documented. (Dave)
b) Ardupilot integrated into an airframe (Jack)
c) GPS onto airframe (Andrew)
d) IR lamp to be bought (Andrew)
e) 2 x 900MHZ 50 mW Xbee ( Andrew )
f) T-shirts / brouchures designed ( Andrew)
See you . .. I'm back next Monday.
Andrew
--
This email remains the property of ACT Department of Education & Training.
This transmission and any accompanying attachments may contain confidential or
legally privileged information. If you are not the intended addressee, you
are notified that any use or dissemination of this email is strictly
forbidden. If you have received this communication in error please notify the
sender immediately and delete all copies of this message. Opinions,
conclusions, views and other information in this message that do not relate to
the official business of ACT Department of Education & Training are the views of the
individual sender and shall be understood as neither given nor endorsed by ACT
Department of Education & Training.
[Canberrauav] Fwd: UAV project status
David Woodgate dpwoodgate at gmail.com
Wed Jan 19 20:30:35 EST 2011
Don't forget Matt, the guy who's already worked on a successful uav, was
going to buy an IR lamp, and is heavily into automatic joe detection because
he's successfully done that type of thing before.
I propose next thursday for the next meeting (wednesday is aus day). My
house if no one else's. 7pm.
---------- Forwarded message ----------
From: "Dave Woodgate"
Date: 19/01/2011 7:46 PM
Subject: Fwd: [Canberrauav] UAV project status
To:
Hi there,
It is good that everyone is back and on-line again (well except for me -
NZ). It is great that the mailing list is up and running as well. There has
been a fair bit of traffic on a number of topics and it would be nice to
get some structure to this information flow. Please feel free to comment on
the following suggestions:
1) We should decide on the team leaders for each of the groups. I suggest :
a) Project Management/requirements
management/deliverables/schedule - Stephen
b) Airframe/autopilot/failsafe - Jack
c) comms - Dave
d) Sponsors/corporate/promotion - Andrew Moss
e) Video processing/target acquisitions - Tidge (Andrew)
f) Ground Station - ??????
g) website - ????
Remember these are the leaders of the groups. They don't have to be the
experts OR do all of the work. There will be a fair bit of interaction
between all of the groups.
2) Do we need separate forums ( mailing lists ) for each group. Is the
information in the Canberra UAV list getting overwhelming already?
3) We should have regular meetings. How about every second Wednesday?
4) Please decide what you want to buy. We are coming to the end of our
"technologies investigations" phase. So far we have bought :
a) 2 x ArduPilots ( integrating
currently )
b) 1 pair of 1mW 2.4 GHZ Xbee radios (
working, range is dismal, very limited usefulness)
c) 1 pair of 50 mW 900Mhz Xbee radios (
working, excellent for medium range testing )
d) 1 x 2.4 GHZ 1 watt amplifier ( not
tested )
e) 1 x 900MHZ 500mWatt video system with
camera and gimbal ( wrong frequency band for Australia, not useable )
f) 1 x fleet of test airframes ( we have
one less now, not sure how many are left )
g) 2 x GPS chips ( Working well, one on
autonomous truck, for software development, 1 metre accuracy, very long
startup, connected to google earth, not tried on plane yet)
h) 1 x AVATAR airframe (for display and
promotional purposes)
i) 3 x Arduino boards ( three different
varieties, all working well, looking at the ECLLIPSE IDE rather than the
Arduino IDE )
5) Website is up and running but really need an web developer.
6) Work to be done over the next week :
a) frequency testing on 900Mhz XBEE. To be documented.
(Dave)
b) Ardupilot integrated into an airframe (Jack)
c) GPS onto airframe (Andrew)
d) IR lamp to be bought (Andrew)
e) 2 x 900MHZ 50 mW Xbee ( Andrew )
f) T-shirts / brouchures designed ( Andrew)
See you . .. I'm back next Monday.
Andrew
--
This email remains the property of ACT Department of Education& Training.
This transmission and any accompanying attachments may contain confidential
or
legally privileged information. If you are not the intended addressee, you
are notified that any use or dissemination of this email is strictly
forbidden. If you have received this communication in error please notify
the
sender immediately and delete all copies of this message. Opinions,
conclusions, views and other information in this message that do not relate
to
the official business of ACT Department of Education& Training are the
views of the
individual sender and shall be understood as neither given nor endorsed by
ACT
Department of Education& Training.
[Canberrauav] Fwd: UAV project status
tridge at samba.org tridge at samba.org
Wed Jan 19 20:41:45 EST 2011
Hi David,
> I propose next thursday for the next meeting (wednesday is aus
> day). My house if no one else's. 7pm.
Unfortunately I'm going to be away until the 3rd of February. You
should still have the meeting I think - I can read about any results
on the list :-)
I'm going to keep working on my FlightGear<->Ardupilot setup and
hopefully get something reasonable working soon after I get back. My
real Ardupilot should have arrived by then.
I've also ordered a 2W 2.4G USB wifi adapter from ebay:
http://cgi.ebay.com.au/ws/eBayISAPI.dll?ViewItem&item=290500695061&ssPageName=STRK:MEWNX:IT
Even if we don't use it, it will be fun to see what sort of range I
can get.
Cheers, Tridge
[Canberrauav] UAV project status
tridge at samba.org tridge at samba.org
Wed Jan 19 20:54:29 EST 2011
Hi Andrew,
> e) Video processing/target acquisitions - Tidge (Andrew)
I'd like to add simulation to my set of tasks, given that is what has
sparked my interest at the moment. I'd like to get a really good
simulator setup going, including:
- "hardware in the loop" simulation (ie. a real arduino running our
actual autopilot code and talking to a flightsim)
- have a real RC connnected to the simulator
- include a real radio link in the simulation
- include simulation of the target acquisition and payload drop
in some form (fake up video link via sat images and target location?)
> Remember these are the leaders of the groups. They don't have to be
> the experts OR do all of the work.
good! I am most definately not an expert on any of these topics. I had
to lookup some of the terms people have been using on wikipedia just
to understand the discussions :-)
> 2) Do we need separate forums ( mailing lists ) for each group. Is
> the information in the Canberra UAV list getting overwhelming
> already?
I'd suggest we not split up the lists - I think there is far too much
interaction between the tasks for that.
People who find the volume too high can set "daily digest" mode in
for their subscription in mailman instead.
> 3) We should have regular meetings. How about every second Wednesday?
sounds good to me. I'd also like to come out to more of the CMAC
meetings at the airfield on the Monaro highway.
Jack, will it be a problem for the CMAC if we regularly have stray UAV
enthusiasts turning up to the Sunday flying sessions, carrying oddball
computer gear?
> 4) Please decide what you want to buy. We are coming to the end of our "technologies
> investigations" phase. So far we have bought :
> a) 2 x ArduPilots ( integrating currently )
I've ordered one as well (a Mega with an IMU board, and associated cables).
> d) 1 x 2.4 GHZ 1 watt amplifier ( not tested )
as mentioned previously, I've ordered a 2W ALFA AWUS036NH 2.4GHz wifi
USB adapter with a 5dBi antenna.
> i) 3 x Arduino boards ( three different varieties, all working well, looking at the ECLLIPSE
> IDE rather than the Arduino IDE )
I have a Arduino TwentyTen as well. That's what I'm using for running
Ardupilot right now.
> 5) Website is up and running but really need an web developer.
I'd suggest we put up a wiki to start with. I could do that if given
admin site access.
Cheers, Tridge
[Canberrauav] FW: Fwd: UAV project status
Stephen Dade stephen_dade at hotmail.com
Wed Jan 19 21:27:52 EST 2011
7pm Thursday is good for me
In response for Andrew:
1) a) I'm happy with this role
f) I'll see if I can interest any friends from Uni
g) My housemate does web design - I should be able to bribe her to create a simple (something
for potential sponsors) site for us
2) The current mailing list works fine in it's current form
3) Every second Wednesday sounds good. Perhaps we would borrow the MHV workspace for the venue?
(I've not seen the workspace yet, so I'm not sure if it's set up for this)
4) I was planning to buy some camera equipment for testing purposes (such as this:
http://www.e-consystems.com/omapovero.asp ), and perhaps a cheap webcam - but I'll defer to Tridge
as team leader. I'm happy to buy/test some of the equipment if he wants.
One point to raise - Has/Is MHV offering us workshop space for our parts/equipment? We're going to
need a common space to work on/store the hardware soon.
5) I know some bare basics of web development. I *may* be able to maintain the site if it's not complicated
and we're not able to find anyone else.
6) g) Contact the UAV challenge organisers to see when they'll be releasing the 2011 rules and if there'll
be any significant changes from 2010 (Stephen)
Thanks
Stephen
Date: Wed, 19 Jan 2011 20:30:35 +1100
From: dpwoodgate at gmail.com
To: canberrauav at canberrauav.com
Subject: [Canberrauav] Fwd: UAV project status
Don't forget Matt, the guy who's already worked on a successful uav, was going to buy an IR lamp, and is
heavily into automatic joe detection because he's successfully done that type of thing before.
I propose next thursday for the next meeting (wednesday is aus day). My house if no one else's. 7pm.
---------- Forwarded message ----------
[Canberrauav] UAV project status
Moss, Andrew Andrew.Moss at ed.act.edu.au
Wed Jan 19 19:11:54 EST 2011
Hi there,
It is good that everyone is back and on-line again (well except for me - NZ). It is
great that the mailing list is up and running as well. There has been a fair bit of traffic
on a number of topics and it would be nice to get some structure to this information
flow. Please feel free to comment on the following suggestions:
1) We should decide on the team leaders for each of the groups. I suggest :
a) Project Management/requirements management/deliverables/schedule - Stephen
b) Airframe/autopilot/failsafe - Jack
c) comms - Dave
d) Sponsors/corporate/promotion - Andrew Moss
e) Video processing/target acquisitions - Tidge (Andrew)
f) Ground Station - ??????
g) website - ????
Remember these are the leaders of the groups. They don't have to be the experts OR do
all of the work. There will be a fair bit of interaction between all of the groups.
2) Do we need separate forums ( mailing lists ) for each group. Is the information in the
Canberra UAV list getting overwhelming already?
3) We should have regular meetings. How about every second Wednesday?
4) Please decide what you want to buy. We are coming to the end of our "technologies
investigations" phase. So far we have bought :
a) 2 x ArduPilots ( integrating currently )
b) 1 pair of 1mW 2.4 GHZ Xbee radios ( working, range is dismal, very limited usefulness)
c) 1 pair of 50 mW 900Mhz Xbee radios ( working, excellent for medium range testing )
d) 1 x 2.4 GHZ 1 watt amplifier ( not tested )
e) 1 x 900MHZ 500mWatt video system with camera and gimbal ( wrong frequency band
for Australia, not useable )
f) 1 x fleet of test airframes ( we have one less now, not sure how many are left )
g) 2 x GPS chips ( Working well, one on autonomous truck, for software development, 1 metre
accuracy, very long startup, connected to google earth, not tried on plane yet)
h) 1 x AVATAR airframe (for display and promotional purposes)
i) 3 x Arduino boards ( three different varieties, all working well, looking at the ECLLIPSE IDE
rather than the Arduino IDE )
5) Website is up and running but really need an web developer.
6) Work to be done over the next week :
a) frequency testing on 900Mhz XBEE. To be documented. (Dave)
b) Ardupilot integrated into an airframe (Jack)
c) GPS onto airframe (Andrew)
d) IR lamp to be bought (Andrew)
e) 2 x 900MHZ 50 mW Xbee ( Andrew )
f) T-shirts / brouchures designed ( Andrew)
See you . .. I'm back next Monday.
Andrew
--
This email remains the property of ACT Department of Education & Training.
This transmission and any accompanying attachments may contain confidential or
legally privileged information. If you are not the intended addressee, you
are notified that any use or dissemination of this email is strictly
forbidden. If you have received this communication in error please notify the
sender immediately and delete all copies of this message. Opinions,
conclusions, views and other information in this message that do not relate to
the official business of ACT Department of Education & Training are the views of the
individual sender and shall be understood as neither given nor endorsed by ACT
Department of Education & Training.
[Canberrauav] Fwd: UAV project status
Dave Woodgate dpwoodgate at gmail.com
Wed Jan 19 22:58:28 EST 2011
On 19/01/2011 8:41 PM, tridge at samba.org wrote:
> Hi David,
>
> > I propose next thursday for the next meeting (wednesday is aus
> > day). My house if no one else's. 7pm.
>
> Unfortunately I'm going to be away until the 3rd of February. You
> should still have the meeting I think - I can read about any results
> on the list :-)
I'd prefer to have tridge available for the meeting. One of the main
purposes of the meeting is to get everybody on the same page as I can
see some people overlapping on some things.
Perhaps we have our first every-second-Wednesday meeting on Thursday the
3rd of Feb (then every 2nd wednesday after that) .
Regarding Steve's suggestion of using the MHV space for the meetings,
I'd have to consult the committee because there are specific rules about
non-members being in the space.
Dave
[Canberrauav] Fwd: UAV project status
Moss, Andrew Andrew.Moss at ed.act.edu.au
Thu Jan 20 17:13:33 EST 2011
Tridge,
Are you buying the flight simulator? I need to buy one for Dickson College. Do you know which
one we need to integrate with the Ardupilot?
Regards,
Andrew
________________________________________
From: canberrauav-bounces at canberrauav.com [canberrauav-bounces at canberrauav.com]
On Behalf Of tridge at samba.org [tridge at samba.org]
Sent: Wednesday, 19 January 2011 8:41 PM
To: David Woodgate
Cc: canberrauav at canberrauav.com
Subject: Re: [Canberrauav] Fwd: UAV project status
Hi David,
> I propose next thursday for the next meeting (wednesday is aus
> day). My house if no one else's. 7pm.
Unfortunately I'm going to be away until the 3rd of February. You
should still have the meeting I think - I can read about any results
on the list :-)
I'm going to keep working on my FlightGear<->Ardupilot setup and
hopefully get something reasonable working soon after I get back. My
real Ardupilot should have arrived by then.
I've also ordered a 2W 2.4G USB wifi adapter from ebay:
http://cgi.ebay.com.au/ws/eBayISAPI.dll?ViewItem&item=290500695061&ssPageName=STRK:MEWNX:IT
Even if we don't use it, it will be fun to see what sort of range I
can get.
Cheers, Tridge
_______________________________________________
Canberrauav mailing list
Canberrauav at canberrauav.com
http://www.canberrauav.com/mailman/listinfo/canberrauav
--
This email remains the property of ACT Department of Education & Training.
This transmission and any accompanying attachments may contain confidential or
legally privileged information. If you are not the intended addressee, you
are notified that any use or dissemination of this email is strictly
forbidden. If you have received this communication in error please notify the
sender immediately and delete all copies of this message. Opinions,
conclusions, views and other information in this message that do not relate to
the official business of ACT Department of Education & Training are the views of the
individual sender and shall be understood as neither given nor endorsed by ACT
Department of Education & Training.
[Canberrauav] Fwd: UAV project status
tridge at samba.org tridge at samba.org
Fri Jan 21 00:08:08 EST 2011
Hi Andrew,
> Are you buying the flight simulator?
I've ordered a ArduPilotMega and associated IMU board and cables.
> I need to buy one for Dickson College. Do you know which one we
> need to integrate with the Ardupilot?
The answer to that is a bit involved ....
I've been working a lot with the Ardupilot code over the last few
days, and I was surprised to find how many varients there are of
it. There are several versions of "Ardupilot" plus several versions of
"ArduPilotMega" and there there are varients such as "MatrixPilot" and
lots of minor branches. These code varients correspond to hardware
varients as well.
The one I've worked the most with this week is "ArduPilot_2_7", which
is a development of the original Ardupilot. I have it running on my
TwentyTen Arduino, and controlling the FlightGear flight simulator
quite nicely. Following a set of waypoints works well, but takeoffs
and landings are quite dodgy so far. I've only had one good automated
landing, most of the time it tends to get very close to landing, then
it gives up and comes around again for another try. I'm glad my
simulator doesn't run out of fuel! Takeoffs are also not good, with
the autopilot too often trying a steep turn as soon as it gets off the
ground, which usually results in a crash.
I had to make some modifications to the autopilot code to get it to
cope with the fact that I don't have a RC control attached to my
simulator, and the fact that my EEPROM values aren't the ones that the
simulator was expecting, so for example I had to disable the mixing
code that tries to combine the trim controls from the RC with the
servo values that Ardupilot is giving the simulator.
After getting all this working, I then started to look at some of the
other ArduPilot varients. I had originally chosen the ArduPilot_2_X
series for my experiments as it seemed easiest to get working on my
TwentyTen board, as it has very little dependence on the IMU shield. I
think that was a mistake however, and looking at the current
ArduPilotMega code I can see that it is much more sophisticated and
also much more flexible.
So to finally answer your question, I suspect we will end up basing
our autopilot on the ArduPilotMega varient of the hardware and
software. That happens to be what I've ordered (by a lucky
coincidence!). This is the page describing what I ordered from
DIYDrones.com:
http://diydrones.com/profiles/blogs/ardupilot-mega-home-page
I got the base ArduPilotMega board, and the Rev-H IMU to go with it. I
haven't ordered a GPS yet (as I didn't think I'd need it for my
simulation work).
It is a bit more expensive than the basic Ardupilot boards, but it
also seems to be a lot more capable. I paid around AU$300 for both
boards plus some cables and shipping to Canberra.
Jack, which Ardupilot varient do you have?
Regardless of which Ardupilot we end up using I think we will need to
make quite a few customisations to it. The code just doesn't do the
type of search pattern we want. For example, it has a single value for
the radius around a waypoint, and if it misses the waypoint it comes
around a few times to try again. We don't want it doing that during
the initial search, but we do want very accurate flight paths for the
package drop.
I also think we'll need careful mods to get the timestamps really
precise for photos (as Matt suggested), and code to task the autopilot
to get photo coverage of a region.
We will also need to be able to change the flightplan and waypoints
mid-flight, which currently Ardupilot can't do. Currently you need to
load a special "waypoint writer" firmware to setup the flight plan. We
don't want to be reloading new firmware mid-flight!
Cheers, Tridge
[Canberrauav] Fwd: UAV project status
Moss, Andrew Andrew.Moss at ed.act.edu.au
Fri Jan 21 05:56:05 EST 2011
Hi Tridge,
Thanks very much for that. I have an Ardupilot on the way. In fact it should be waiting at my
door when I return on Monday.
I will get a flightgear flight simulator next week as well.
Thanks again,
Andrew
________________________________________
From: tridge at samba.org [tridge at samba.org]
Sent: Friday, 21 January 2011 12:08 AM
To: Moss, Andrew
Cc: canberrauav at canberrauav.com
Subject: Re: [Canberrauav] Fwd: UAV project status
Hi Andrew,
> Are you buying the flight simulator?
I've ordered a ArduPilotMega and associated IMU board and cables.
> I need to buy one for Dickson College. Do you know which one we
> need to integrate with the Ardupilot?
The answer to that is a bit involved ....
I've been working a lot with the Ardupilot code over the last few
days, and I was surprised to find how many varients there are of
it. There are several versions of "Ardupilot" plus several versions of
"ArduPilotMega" and there there are varients such as "MatrixPilot" and
lots of minor branches. These code varients correspond to hardware
varients as well.
The one I've worked the most with this week is "ArduPilot_2_7", which
is a development of the original Ardupilot. I have it running on my
TwentyTen Arduino, and controlling the FlightGear flight simulator
quite nicely. Following a set of waypoints works well, but takeoffs
and landings are quite dodgy so far. I've only had one good automated
landing, most of the time it tends to get very close to landing, then
it gives up and comes around again for another try. I'm glad my
simulator doesn't run out of fuel! Takeoffs are also not good, with
the autopilot too often trying a steep turn as soon as it gets off the
ground, which usually results in a crash.
I had to make some modifications to the autopilot code to get it to
cope with the fact that I don't have a RC control attached to my
simulator, and the fact that my EEPROM values aren't the ones that the
simulator was expecting, so for example I had to disable the mixing
code that tries to combine the trim controls from the RC with the
servo values that Ardupilot is giving the simulator.
After getting all this working, I then started to look at some of the
other ArduPilot varients. I had originally chosen the ArduPilot_2_X
series for my experiments as it seemed easiest to get working on my
TwentyTen board, as it has very little dependence on the IMU shield. I
think that was a mistake however, and looking at the current
ArduPilotMega code I can see that it is much more sophisticated and
also much more flexible.
So to finally answer your question, I suspect we will end up basing
our autopilot on the ArduPilotMega varient of the hardware and
software. That happens to be what I've ordered (by a lucky
coincidence!). This is the page describing what I ordered from
DIYDrones.com:
http://diydrones.com/profiles/blogs/ardupilot-mega-home-page
I got the base ArduPilotMega board, and the Rev-H IMU to go with it. I
haven't ordered a GPS yet (as I didn't think I'd need it for my
simulation work).
It is a bit more expensive than the basic Ardupilot boards, but it
also seems to be a lot more capable. I paid around AU$300 for both
boards plus some cables and shipping to Canberra.
Jack, which Ardupilot varient do you have?
Regardless of which Ardupilot we end up using I think we will need to
make quite a few customisations to it. The code just doesn't do the
type of search pattern we want. For example, it has a single value for
the radius around a waypoint, and if it misses the waypoint it comes
around a few times to try again. We don't want it doing that during
the initial search, but we do want very accurate flight paths for the
package drop.
I also think we'll need careful mods to get the timestamps really
precise for photos (as Matt suggested), and code to task the autopilot
to get photo coverage of a region.
We will also need to be able to change the flightplan and waypoints
mid-flight, which currently Ardupilot can't do. Currently you need to
load a special "waypoint writer" firmware to setup the flight plan. We
don't want to be reloading new firmware mid-flight!
Cheers, Tridge
--
This email remains the property of ACT Department of Education & Training.
This transmission and any accompanying attachments may contain confidential or
legally privileged information. If you are not the intended addressee, you
are notified that any use or dissemination of this email is strictly
forbidden. If you have received this communication in error please notify the
sender immediately and delete all copies of this message. Opinions,
conclusions, views and other information in this message that do not relate to
the official business of ACT Department of Education & Training are the views of the
individual sender and shall be understood as neither given nor endorsed by ACT
Department of Education & Training.
[Canberrauav] Fwd: UAV project status
tridge at samba.org tridge at samba.org
Fri Jan 21 08:42:19 EST 2011
Hi Andrew,
> > Are you buying the flight simulator?
oh, I just realised, you were asking about buying the flight
simulator, somehow I read it as buying the autopilot! Serves me right
for doing email at midnight.
The FlightGear flight simulator is free, and comes with source code.
See http://flightgear.org/. I'm using the 2.0.0 version.
It is a very good simulator, and has very flexible interfaces for
connecting it to real hardware like an ArduPilot.
Cheers, Tridge
[Canberrauav] Fwd: UAV project status
Dave Woodgate dpwoodgate at gmail.com
Fri Jan 21 08:39:08 EST 2011
On 21/01/2011 12:08 AM, tridge at samba.org wrote:
> Hi Andrew,
>
> > Are you buying the flight simulator?
>
> I've ordered a ArduPilotMega and associated IMU board and cables.
>
> > I need to buy one for Dickson College. Do you know which one we
> > need to integrate with the Ardupilot?
>
> The answer to that is a bit involved ....
>
> I've been working a lot with the Ardupilot code over the last few
> days, and I was surprised to find how many varients there are of
> it. There are several versions of "Ardupilot" plus several versions of
> "ArduPilotMega" and there there are varients such as "MatrixPilot" and
> lots of minor branches. These code varients correspond to hardware
> varients as well.
>
> The one I've worked the most with this week is "ArduPilot_2_7", which
> is a development of the original Ardupilot. I have it running on my
> TwentyTen Arduino, and controlling the FlightGear flight simulator
> quite nicely. Following a set of waypoints works well, but takeoffs
> and landings are quite dodgy so far. I've only had one good automated
> landing, most of the time it tends to get very close to landing, then
> it gives up and comes around again for another try. I'm glad my
> simulator doesn't run out of fuel! Takeoffs are also not good, with
> the autopilot too often trying a steep turn as soon as it gets off the
> ground, which usually results in a crash.
Dude. This simulation you're doing is churning out some quite valuable
information about auto-pilot behavior under certain circumstances.
Without risking the real UAV! Well done.
> ....
>
> Cheers, Tridge
> _______________________________________________
> Canberrauav mailing list
> Canberrauav at canberrauav.com
> http://www.canberrauav.com/mailman/listinfo/canberrauav
>
[Canberrauav] Fwd: UAV project status
tridge at samba.org tridge at samba.org
Fri Jan 21 10:00:44 EST 2011
Hi Dave,
> Dude. This simulation you're doing is churning out some quite valuable
> information about auto-pilot behavior under certain circumstances.
> Without risking the real UAV! Well done.
thanks!
My ArduPilotMega board has now arrived, so I just need to get it
soldered up and I should be able to build a simulation setup that much
more accurately reflects what we would use on the real plane.
Cheers, Tridge
[Canberrauav] Fwd: UAV project status
tridge at samba.org tridge at samba.org
Fri Jan 21 10:17:24 EST 2011
Hi Dave,
Some more comments on why I think simulation is important.
We're going to have to do lots of real flight testing, but given the
timeframe we won't have the luxury of being able to test in all the
conditions we may encounter on the day.
For example, will the autopilot cope with a 20 knot wind? Will it end
up going outside the mission boundary, or know that it needs to
account for the wind when planning the course for the package drop?
With a simulation we can setup that wind, and see if it copes. I
suspect we'll find that the autopilot doesn't cope with it at all
well, and we'll need to modify the code to better handle wind,
especially if it is gusty. The simulation won't be a perfact match for
the real model in wind, but it will at least give us some idea of
whether the autopilot can cope at all.
I'll put together a demo of the simulation for the group at one of our
evening meetings.
Jack, can you tell me what flight control channels we're likely to end
up with on the plane? Right now I'm using a model of a SIG Rascal 110
RC plane. It's basic metrics are:
wingspan: 2.8m
wingarea: 0.98m^2
controls:
- elevator
- single linked aileron control
- rudder
- throttle
See more details here: http://www.sigmfg.com/IndexText/SIGRC84ARFB.html
Is that a reasonable match for what you think we'll use?
Cheers, Tridge
Copyright 2011 http://canberrauav.org.au/