Project Universal Autopilots alpha 100324

Status
Not open for further replies.

Artlav

Aperiodic traveller
Addon Developer
Beta Tester
Joined
Jan 7, 2008
Messages
5,672
Reaction score
453
Points
173
Location
Earth
Website
orbides.org
Alpha 100317:
orbides.1gb.ru/orbf/uap-alpha-100317.zip
VC6 build (if above not working): http://orbides.1gb.ru/orbf/uap-alpha-100317-vc6.zip

Changes:
-Fixed multiple orientation mishaps (retrograde, many APs in high-eccentricity orbit, sync, etc)
-Fixed keys again
-Fixed multi-pass align planes thrust leak
-Fixed waiting bug
-Tuned time accel controls to buffer large time accelerations (align, hohmann)
-Integrated UAP into Spaceway
-Added Hohmann autopilot
-Added triggering other vessels to the tools
-Added lift/pitch (main+hover) to lift off AP

Added autopilots:
Hohmann autopilot
Gets you from an orbit of a planet to an orbit of its moon
- engine specifies the engine to use
- target specifies the target moon/planet to go to
- tgt_orbit_alt defines the altitude of the orbit around destination

Scenarios:
-Hohmann-DG KSC to the Moon station - Gets a DG from Earth to a station in lunar orbit.
-Martian chase - one SH-A chases the other to orbit, including all manoeuvres in local space (by mjanicki)
-Mission-SH-A and DG to Luna OB1 - Shuttle A and DG race from Brighton to Luna-OB1 in different trajectories.
-Launch-Many from the Moon - one of the DGs now launches in Slideshow's proposed orientation.

Hohmann transfer, a first interplanetary tool, is the focus of this alpha, along with bug fixes. This AP will only take a ship from a planet to it's Moon, but expanding it towards a full interplanetary flight is a small step.

Waiting is the only problem - launch windows between planets can be years apart. Is it worth making, or better do a parametric transfer one instead?
 
Last edited:

KosmoKen

Lost in Translation
Donator
Joined
Feb 4, 2009
Messages
84
Reaction score
0
Points
0
Location
Tucson
Hohmann autopilot is very nice. Is it yet possible to choose a target inclination, and if so a retrograde one? Or is that yet to be implemented?

Also I noticed that engine thrust on approach and orbit insert is done in many short, full throttle, spurts. It works nice but is ugly. To be changed?
 

Artlav

Aperiodic traveller
Addon Developer
Beta Tester
Joined
Jan 7, 2008
Messages
5,672
Reaction score
453
Points
173
Location
Earth
Website
orbides.org
Hohmann autopilot is very nice. Is it yet possible to choose a target inclination, and if so a retrograde one? Or is that yet to be implemented?
I'm planning to split it into transfer part - hohmann or whatever else will there be to send a vessel on an interplanetary trajectory and navigate it to the target, and arrival, to guide it from an approach hyperbola into an orbit of given parameters.

Also I noticed that engine thrust on approach and orbit insert is done in many short, full throttle, spurts. It works nice but is ugly. To be changed?
Fixed already.
 

Star Voyager

Space Shuttle Refugee
Joined
Oct 25, 2008
Messages
1,973
Reaction score
29
Points
48
Alpha 100317:
http://orbides.1gb.ru/orbf/uap-alpha-100317.zip

Changes:
-Fixed multiple orientation mishaps (retrograde, many APs in high-eccentricity orbit, sync, etc)
-Fixed keys again
-Fixed multi-pass align planes thrust leak
-Fixed waiting bug
-Tuned time accel controls to buffer large time accelerations (align, hohmann)
-Integrated UAP into Spaceway
-Added Hohmann autopilot
-Added triggering other vessels to the tools
-Added lift/pitch (main+hover) to lift off AP

Added autopilots:
Hohmann autopilot
Gets you from an orbit of a planet to an orbit of its moon
- engine specifies the engine to use
- target specifies the target moon/planet to go to
- tgt_orbit_alt defines the altitude of the orbit around destination

Scenarios:
-Hohmann-DG KSC to the Moon station - Gets a DG from Earth to a station in lunar orbit.
-Martian chase - one SH-A chases the other to orbit, including all manoeuvres in local space (by mjanicki)
-Mission-SH-A and DG to Luna OB1 - Shuttle A and DG race from Brighton to Luna-OB1 in different trajectories.
-Launch-Many from the Moon - one of the DGs now launches in Slideshow's proposed orientation.

Hohmann transfer, a first interplanetary tool, is the focus of this alpha, along with bug fixes. This AP will only take a ship from a planet to it's Moon, but expanding it towards a full interplanetary flight is a small step.

Waiting is the only problem - launch windows between planets can be years apart. Is it worth making, or better do a parametric transfer one instead?

Any chance compliling this with VC6? I'm still using UAP 101304-VC6 and that's the only way it works on my computer.
 

Artlav

Aperiodic traveller
Addon Developer
Beta Tester
Joined
Jan 7, 2008
Messages
5,672
Reaction score
453
Points
173
Location
Earth
Website
orbides.org

Sideshow

New member
Joined
Jul 1, 2009
Messages
31
Reaction score
0
Points
0
-Added lift/pitch (main+hover) to lift off AP

Just tried it and it worked great, exactly what I'd envisaged. Thank you :thumbup:

The Hohmann autopilot worked very well using the demo scenario, putting the DG into lunar orbit without issue even with me using up to 1000x time compression for some of the trip. Unfortunately the align planes autopilot would not align me with the station so thats as far as I got. About to try again.

---edit---

Ok with a little assistance during the align planes maneuver with the station all works ok. Now trying out different target altitudes.

I have another question though. When launching a UMMU 2.0 ShuttleA from KSC, an initial launch heading of 42 degrees and using the following autopilot sequences everything goes well until the trans_orbit sequences. Upon starting this sequence the vessel immediately yaws from 70 degrees to about 78 degrees. Is this intentional? I thought it would continue on its last heading.

Code:
  lift_off(engine=main,mode=3,target=,heading=70.000000,altitude=500,pitch_tgt=80.000000,pitch_duration=60.000000,off_duration=5.000000,stage=0), 
  tools(type=0,key=g),
  lift_off(engine=main,mode=2,heading=70,pitch_tgt=50,pitch_duration=60,off_duration=0,stage=0),
  trans_orbit(engine=main,kind=0,apoapsis=200000,periapsis=200000,mode=0).

--- edit---

Further playing around showed that when it engages the trans_orbit sequence autopilot it yaws to align the vessel prograde. In my example 70 degrees was not exactly prograde but 78 was (confirmed by switching the HUD to orbit mode instead of surface mode). So its not a bug or problem with the autopilot just with the person using it :lol:
 
Last edited:

jgrillo2002

Conservative Pioneer
Addon Developer
Joined
Mar 17, 2008
Messages
755
Reaction score
16
Points
33
Location
New York State
Same problem. I think the align plane autopilot only works if the vessel is going to a prograde orbit. that needs to be fixed
 

Artlav

Aperiodic traveller
Addon Developer
Beta Tester
Joined
Jan 7, 2008
Messages
5,672
Reaction score
453
Points
173
Location
Earth
Website
orbides.org
Alpha 100318:
orbides.1gb.ru/orbf/uap-alpha-100318.zip
In post 94

Changes:
-Added thrust reduction to runway take-off, halves fuel wasted on take-off
-Fixed targetting celestial bodies in all AP's where they can be targeted
-Fixed intermitten thrust in hohmann
-Fixed 90 degrees limit in align planes, any angle is good now
-Fixed several cases of eccentricity in hohmann
-Improved hohmann accuracy

Scenarios:
-Hohmann-DG KSC to the Moon station - Gets a DG from Earth to a station in lunar orbit, now without any assistance

Unfortunately the align planes autopilot would not align me with the station so thats as far as I got.
That should now be fixed.

I have another question though. When launching a UMMU 2.0 ShuttleA from KSC, an initial launch heading of 42 degrees and using the following autopilot sequences everything goes well until the trans_orbit sequences. Upon starting this sequence the vessel immediately yaws from 70 degrees to about 78 degrees. Is this intentional?
Yes, it launches into the current plane if no target is given.
 
Last edited:

mjanicki

Addon Developer
Addon Developer
Donator
Joined
Apr 4, 2008
Messages
106
Reaction score
0
Points
0
Location
Spokane
Code:
  lift_off(engine=main,mode=3,target=,heading=70.000000,altitude=500,pitch_tgt=80.000000,pitch_duration=60.000000,off_duration=5.000000,stage=0), 
  tools(type=0,key=g),
  lift_off(engine=main,mode=2,heading=70,pitch_tgt=50,pitch_duration=60,off_duration=0,stage=0),
  trans_orbit(engine=main,kind=0,apoapsis=200000,periapsis=200000,mode=0).
--- edit---

Further playing around showed that when it engages the trans_orbit sequence autopilot it yaws to align the vessel prograde. In my example 70 degrees was not exactly prograde but 78 was (confirmed by switching the HUD to orbit mode instead of surface mode). So its not a bug or problem with the autopilot just with the person using it :lol:

Nothing wrong with that, Sideshow. I fought with the exact same thing last night with my ShuttleA and settled on the same sort of program for the autopilot to use the new mode 3 lift off. The vessel was definitely yawing to point in line with its velocity vector so the second lift_off burn helps by getting the vessel moving in the right direction.

I noticed that having a vessel with no control surfaces in an atmosphere contributes to the yaw as well. A ShuttleA on the moon can do well going from mode 3 liftoff straight into a trans_orbit, but on Mars the vessel will yaw back and forth fighting the atmosphere. Including an off_duration in the second lift_off can help with that as well. On Mars I used off_duration=30 which allowed the vessel to both increase forward velocity and reach a thinner part of the atmosphere. I was able to keep the vessel moving forward with trans_orbit that way.

Anyway... I was typing up a long post about the yaw and other things I'd tested; but Artlav uploaded 100318 while I was typing. So, I'll save the rest of the comments and go test the new release.

-- Mike
 

IronRain

The One and Only (AFAIK)
Administrator
Moderator
News Reporter
Donator
Joined
Oct 11, 2009
Messages
3,476
Reaction score
394
Points
123
Location
Utrecht
Website
www.spaceflightnewsapi.net
a small question.

UAP isn't compateble with Universal Crago deck, is it?
because I tried it but it didn't brought me to a good orbit (actualy, I crashed back to earth ;))
 

Artlav

Aperiodic traveller
Addon Developer
Beta Tester
Joined
Jan 7, 2008
Messages
5,672
Reaction score
453
Points
173
Location
Earth
Website
orbides.org
UAP isn't compateble with Universal Crago deck, is it?
because I tried it but it didn't brought me to a good orbit (actualy, I crashed back to earth ;))
Hm?
What exactly have you tried?
Cargo deck is just an attachment manager, shouldn't interfere with guidance, unless it skews the mass or something.
 

IronRain

The One and Only (AFAIK)
Administrator
Moderator
News Reporter
Donator
Joined
Oct 11, 2009
Messages
3,476
Reaction score
394
Points
123
Location
Utrecht
Website
www.spaceflightnewsapi.net
Hm?
What exactly have you tried?
Cargo deck is just an attachment manager, shouldn't interfere with guidance, unless it skews the mass or something.

well, I've added the Carina satellite to the jarvis-H rocket with Universal Cargo Deck and then launched using UAP 100314. then I got in a realy unstable orbit and eventually crashed back to earth
 

mjanicki

Addon Developer
Addon Developer
Donator
Joined
Apr 4, 2008
Messages
106
Reaction score
0
Points
0
Location
Spokane
alpha-100318 feedback

A few things that were problematic in 100317 seem to have been addressed in 100318, so all looking good. I didn't do extensive testing, but I did run through all my usual testing scenarios.

Vertical launch with a multistage vessel looked good. With an Atlas V there is still a bit of rolling back and forth during the initial phase of the launch, but I'm convinced that's my problem since no other vessel seems to roll like this.

Align planes is back to being very accurate. Problems that I still saw in 100317 seem to have been fixed in 100318. Nice job.

Sync Orbit seemed more accurate than in previous releases and worked well.

The autopilots handling of time acceleration seems to be better in this release. It might just be my imagination, but it looked like the autopilots were more sensible about deciding when to reduce time acceleration to allow for a maneuver.

I have noticed that each release seems to come with more information displayed in the MFD. The useful "tips" in each of the main pages are a very nice addition.

Thank you for adding the mode=3 lift_off! This is exactly what a ShuttleA needs. I've converted all my ShuttleA tests to use this mode and it works perfectly, provided I remember to put the engine type as "main" and not "hover."

One problem with 100318 -- targeted lift_off and trans_orbit don't seem to be targeting when the target is a vessel. It was working up until 100317, but not in 100318. You can see it even in the demo martian chase scenario.

I did test one Hohmann transfer with a quick run from Earth to Moon using Deepstar -- partly just to test some other vessels, partly because I had it sitting in orbit already. The transfer went well and a problem with the informational display I saw in 100317 seems to have been fixed in 100318. One potential difficulty with the Hohmann AP was that it began the Moon insertion burn much earlier than I'd expected. As a result, the final PeA was 150 km lower than had been programmed (600 km), for a 25% error from the programmed altitude. But the trajectory and corrections were good and the insertion was successful overall.

That's about it, except for one wish-list item. Could we get the "e" key added to the keypress tool? That would allow me to program the EVA of a UMMU once a vessel has reached orbit.

Thanks again for all the hard work, Artlav. :thumbup:

-- Mike
 

Artlav

Aperiodic traveller
Addon Developer
Beta Tester
Joined
Jan 7, 2008
Messages
5,672
Reaction score
453
Points
173
Location
Earth
Website
orbides.org
Alpha 100318-1:
orbides.1gb.ru/orbf/uap-alpha-100318-1.zip

Changes:
-Fixed a screw-up with targetting in all autopilots.
-Added E to key table

One problem with 100318 -- targeted lift_off and trans_orbit don't seem to be targeting when the target is a vessel. It was working up until 100317, but not in 100318. You can see it even in the demo martian chase scenario.
Oops.
 
Last edited:

Pablo49

Member
Joined
Feb 24, 2010
Messages
753
Reaction score
0
Points
16
Wow, I'm so far behind on this alpha now. You sir, are a machine.

EDIT: I finally got the chance to test.

I took a stock DG from Ascension Island, to docking to a station I've been building, then took it to the moon. runway_off to docking worked fine. (Aside from lift off and trans orbit still always assuming a north-eastern launch, the South-Eastern launch failed, using targets. North-Eastern worked fine) I undocked manually, then aligned planes with the moon, which worked fine.

Hohmann transfer worked well, but like mjanicki, it put me in a far lower orbit than I had set. I had a 200 x 200 orbit set, and it put me at about 2 x 140 orbit. I tried to correct with tran_orbit, but it gave me only one short burn then turned off, which put me into a 30 x 160 orbit. (I assume trans_orbit failed because of the odd orbit since I've never had any problems before.) After deorbitting manually, I landed fine with get_on_pad.

This was the first time I've used a DG with UAP, and it got me thinking. Is it possible to add Scram jet settings into lift off and runway off? I personally rarely fly any ships with scram jets, so I barely understand how best to use them. But I know is is more efficient to use them when you have them. Would it be possible to have a setting to use scram between two altitudes, or between two velocities, or something like that?

Also, three minor things: In the Doc, it still says align is only good within 90 degrees, and mode 3 isn't listed in tools, so that needs updating. And why does tools have -1 set as the mode default? That's not a valid input. (But "mode_is: WTF?" is very amusing nonetheless. :lol:)
 
Last edited:

johndeere5

New member
Joined
Mar 16, 2010
Messages
30
Reaction score
0
Points
0
Keep up the good work, now I will have something to use to get to where I'm going, within Earth-Moon to say the least, no matter what I do with the TransX, even follow the manual, all manual and other related close to it, I can't go anywhere, not even to the moon or plane change, it just say "No Command" on the nav screen, had the Redshift before and works perfectly for my need at least to ISS and Moon but not return so its a one-way-ride, but I lost it due to backup drive failure and redownload it again, but it's even worst than before, can't go anywhere, no ISS and Moon selection on list but it's in FOP file and can't access it on the ship nav screen, until I heard about this Universal Autopilot, which shows promising to me, now I'm going to download it to see how it does, can't wait for the final complete version, only thing to say now is keep up the good work and this shows good sings to others that may have difficulty as I have using and/or understanding the TransX. :thumbup::tiphat::cheers::hail:
 

KosmoKen

Lost in Translation
Donator
Joined
Feb 4, 2009
Messages
84
Reaction score
0
Points
0
Location
Tucson
Small request. Any way for "Heading = X " to carry over into next step? Just so I don't have to worry about whether I kept it consistent through the various steps.

Also, is "Engine = 15 " the only way to get a SCRAM engine going through Aerial Hold? It works well enough if that's the case, just curious.\

Another thing: When in Aerial Hold AP, how can I change "alt_rate" without raising "altitude"? I'm trying to get a good Aerial Hold AP sequence for the XR2 Ravenstar's SCRAM cruise phase. For example; have a final "altitude" of 70000 with an "alt_rate" of 100 m/s, with an "engine" of 15(see above), and "velocity" of 6300 in an Aerial Hold sequence. Am I just doing it wrong because I don't completely understand the "onoff"?

One more thing: I don't know if its because I'm doing something wrong but I'm pretty sure every time I open a saved scenario with a UAP sequence, it automatically runs the AP when the scenario starts. Very annoying when I'm trying to tweak a UAP sequence for a specific vessel/orbit.
 
Last edited:

Sideshow

New member
Joined
Jul 1, 2009
Messages
31
Reaction score
0
Points
0
Using alpha 100318-1 I've been furthering my experimentation with with launching the Shuttle A from KSC.

At the moment it is not possible to launch into a specific orbital plane unless there is already a vessel orbiting in that plane for the autopilot to use as a target. My testing has consisted of the following:

1) Launch the ShuttleA from KSC using the following autopilot. I have placed the vessel on the pad heading 000 so that the autopilot does not have to yaw the vessel. The intention is to launch into a polar orbit.
Code:
lift_off(engine=main,mode=3,target=,heading=0.000000,altitude=400.000000,pitch_tgt=80.000000,pitch_duration=60.000000,off_duration=5.000000,stage=0),
tools(type=0,key=g),
lift_off(engine=main,mode=2,heading=0.000000,pitch_tgt=50.000000,pitch_duration=60.000000,off_duration=0,stage=0),
trans_orbit(engine=main,apoapsis=200000.000000,periapsis=200000.000000).

2) When the autopilot gets to the trans_orbit sequence the ShuttleA will yaw to the right to 030 degrees. This is to align the vessel with its velocity vector, which includes my initial velocity from the rotation of the earth, and so is not actually 000 but instead 030. I believe the effect is not as noticeable in the DG since its aerodynamic surfaces are much more effective the the ShuttleA RCS in the lower atmosphere.

There are two possible ways to overcome this problem I think.

1) Make the trans_orbit autopilot check the previous sequence. If it was lift_off then and no target has been specified then do not yaw the vessel to align with the velocity vector. This may have unexpected effect on some autopilot sequences.

2) Put an additional parameter in the trans_orbit autopilot such as "launch = 1" that way the trans_orbit autopilot will know that the user intends to launch into a specific plane not the one it is presently in and will not attempt to yaw the vessel away from that plane.

Thanks for all your efforts Artlav, I may appear a little over enthusiastic but this is something I have been dreaming for in orbiter for quite some time.
 
Last edited:
Status
Not open for further replies.
Top