[WIN/MAC/LINUX] KSP Trajectory Optimization Tool v1.6.10 [Major LVD Improvements!]

[Arrowstar]

And here's KSP TOT v1.2.0 prerelease 7. This has all the fixes I mentioned above for Gaiiden. Still working the longitude issue.

[Drew Kerman]

See if you can spot what was causing the latter

I'm going to say ... deg2rad? Heh, victim of copy/paste?

I figured out what was going on with the Minmus case you sent me... kinda. I'm not sure why, but basically it looks like the spacecraft was started inside Minmus (the body itself) but with the starting SoI set as Kerbin. This produced some wierdness that ultimately made MATLAB try to form an array a few hundred billion rows long. I've added a check to prevent that from happening in the future (max number of states per propagation limited to 10000, normally it doesn't exceed 1000).

Huh, well I did of course load a modified bodies.ini file with Minmus deleted when I first created the mission file to model Minmus' orbit so I wouldn't be crashing into it constantly, but since I learned that orbital period change I made had been saved to the Mun mission file I had you look at earlier I figured I didn't need to have KSPTOT point to the bodies.ini without Minmus every time I loaded that mission file.

And no, I'd prefer to keep the edit event dialogs model. I know it's a pain, but if the mission script gets out of sync with the dialog box, serious problems can ensue. Better safe than sorry.

Good point.

Edited February 3, 2015 by Gaiiden

[Arrowstar]

So I looked into the longitude problem... and I didn't find a problem. Here's what I did:

1) Started KSP and loaded a vessel. Opened KER.

2) Quicksaved KSP and immediately paused the game.

3) Opened MA and imported the orbit from SFS file.

4) Compared the two longitudes. They're only a few hundredths of a degree off.

See:

[Drew Kerman]

So I looked into the longitude problem... and I didn't find a problem

You are correct, there isn't a problem in KSPTOT - there's a problem in my SFS file... or in some way that KSPTOT is reading my SFS file. I posted details in the support forum hoping someone could confirm sooner than you would probably be available again.

[Arrowstar]

I'm pleased to announce that KSP TOT v1.2.0 is now formally released after an extended period of pre-release testing!

WARNING: THIS UPDATE REQUIRED A NEW VERSION OF THE MATLAB COMPILER RUNTIME, VERSION R2014B X64.

New features include:

• Upgraded MATLAB Compiler Runtime to version R2014b x64.
  
• Added a search parameter slider to Rendezvous Maneuver Sequencer that allows the user to adjust the way the optimizer works.
  
• Added ability to split impulsive DV maneuvers in Mission Architect
  
• Changed aerobraking models away from the impulsive approximation and replaced those with numerical integration that should be much more accurate.
  
• Added a function to Mission Architect Graphical Analysis that allows the user to draw lines at specified points on the plots
  
• Lots of other bug fixes and minor tweaks to Mission Architect.
  

Please let me know if you have any questions. Thanks!

Edited February 4, 2015 by Arrowstar

[Sigma88]

this looks interesting, is it possible to limit the options in the flyby (gravity assist) mode?

example:

I want to get to jool from kerbin, launching after year 1 day 250 and arriving before day 550. can this tool calculate which would be the best dV in that interval?

edit:

I noticed now some intervals can be set, it's not exactly what I was thinking of but it'll do

looks like in my current save the planets are not alligned for doing anything interesting in the short term :\

Edited February 4, 2015 by Sigma88

[Arrowstar]

Glad you hear you have things more or less sorted. Apologies about the planetary alignments, maybe use some limited timewarp to get you somewhere better?

[Drew Kerman]

Here's my upcoming Mun mission launching next week!

[Arrowstar]

That is wicked cool.

[Drew Kerman]

thanks. Surprised to see it fall completely flat on reddit tho. You'd think people would be interested in something new posted there for a change

[Drew Kerman]

So I looked into the longitude problem... and I didn't find a problem

I just had an epiphany at 6am while trying to go to sleep. It was driven by what Diomeda said in my support forum thread on this issue, which has been itching my brain for days. Do me a favor and please run your test again, but set the UT of the game to 2yrs instead of 6 days. Or even 4yrs. I'm sure you see where I'm going with this.

Edit: dammit I still couldn't sleep so I tried it myself. Let's just say I'm heading to bed with a much more restful mind now (although still wishing I didn't have to wake up in 4 hours...)

Edited February 7, 2015 by Gaiiden

[GrihanOs]

Hi,

I've been playing with this tool, I like it a lot

I tried to use the multi-flyby maneuver sequencer to plan a mission but there doesn't seem to be a way to time the maneuver to leave Kerbin (origin planet).

There is the burn true anomaly, but with the calculation giving the maneuvers years ahead, this isn't enough.

Say I want to plan a maneuver in the mission planner, the first event should be "Coast to UT" to some time before the maneuver, than an optimizable "Coast to true anomaly", etc.

I can't find the value I put in the "Coast to UT" event, the only time given is "Kerbin Departure Date" which is, to my understanding, the time of SOI change.

What am I missing?

[Arrowstar]

I just had an epiphany at 6am while trying to go to sleep. It was driven by what Diomeda said in my support forum thread on this issue, which has been itching my brain for days. Do me a favor and please run your test again, but set the UT of the game to 2yrs instead of 6 days. Or even 4yrs. I'm sure you see where I'm going with this.

Edit: dammit I still couldn't sleep so I tried it myself. Let's just say I'm heading to bed with a much more restful mind now (although still wishing I didn't have to wake up in 4 hours...)

Brilliant. You just found the problem. I created a scenario 1000 days out and looked at the longitude difference between KER and KSPTOT. It was about 5.8 degrees, too much by far. So I looked into the rotation period for Kerbin and lo, there was the issue.

Here's what I have now:

rotperiod = 21600.0

Here's what KSP reports the value to be (to 6 decimal places):

rotperiod = 21599.912015

Using the latter value for rotperiod in my bodies.ini file for Kerbin yields a longitude error measurable in the hundredths of a degree after 1000 24-hour periods. I think this is probably close enough.

To fix, just update your bodies.ini with the latter value and away you go. I can correct any MA saved missions (MAT files) directly with MATLAB if you want.

Hi,

I've been playing with this tool, I like it a lot

I tried to use the multi-flyby maneuver sequencer to plan a mission but there doesn't seem to be a way to time the maneuver to leave Kerbin (origin planet).

There is the burn true anomaly, but with the calculation giving the maneuvers years ahead, this isn't enough.

Say I want to plan a maneuver in the mission planner, the first event should be "Coast to UT" to some time before the maneuver, than an optimizable "Coast to true anomaly", etc.

I can't find the value I put in the "Coast to UT" event, the only time given is "Kerbin Departure Date" which is, to my understanding, the time of SOI change.

What am I missing?

So the flyby maneuver sequencer (FMS) shouldn't be seen as anything more than a coarse way to find out roughly when to leave and when to arrive somewhere to hit the flyby. If you want to use that output in Mission Architect, I'd recommend setting your initial state epoch a few days prior to whenever FMS told you to leave, and use the delta-v information to create a burn. The True Anomaly of the burn in FMS is actually going to be pretty accurate (plus/minus 5-10 degrees), so just coast to that true anomaly and add revs to the coast if needed. Then let the mission optimizer adjust the true anomaly and burn delta-v components to actually arrive at the planet/moon you're looking to flyby.

Does that help any? Anything still unclear?

[Drew Kerman]

To fix, just update your bodies.ini with the latter value and away you go. I can correct any MA saved missions (MAT files) directly with MATLAB if you want.

And away I went! No need for any MAT file patching. Ahhh this feels good. I hate to admit it but that 2deg difference was really bugging me lol

[Arrowstar]

No worries, glad we got to the bottom of it!

[Drew Kerman]

And no, I'd prefer to keep the edit event dialogs model. I know it's a pain, but if the mission script gets out of sync with the dialog box, serious problems can ensue. Better safe than sorry.

You weren't kidding. If I have a mission script loaded and got to change the UT on a coast event, I found I can accidentally click closed the Edit State window with the X button and then when I close the UT Date/Time window the whole program just goes kaput and kills itself

Also, what is the green diamond on this transfer plot above where I meet up with Eve?

Also, also - Arrowstar do you have any general advice to share about planning with the FMS? Obviously there's no real defined process like there would be with a point-to-point mission, but surely there are some things one should not do when setting up runs with the FMS to avoid repeated "unfeasible results"? I'm just not sure what are some good parameters to define from the start and then how best to tweak things to work towards a result that is worth transferring over to MA to play around with. After 5 runs with "solution may not be feasible" I'm left wondering - well is this just impossible altogether regardless of the constraints I set or do I just need to change these constraints a little to make it work? And how?

I'm also keen on plotting gravity assist missions in the future as well, but my brain kinda melts a little bit when I consider how to go about doing it.

And more stuff - is it possible to display the pop-up tooltip with extended orbital data for the Other Craft window?

I'm also having trouble splitting a burn. I have a burn that will go through two stages with different engines, so I used the burn split tool and gave it the Dv I have remaining in my first stage for the first split and it auto-calculated the second split. I hit save and my end state went from 959km Pe to 1139km Pe. Why is that?

Edited February 9, 2015 by Gaiiden

[Arrowstar]

You weren't kidding. If I have a mission script loaded and got to change the UT on a coast event, I found I can accidentally click closed the Edit State window with the X button and then when I close the UT Date/Time window the whole program just goes kaput and kills itself

Can you tell me how to reproduce this? Sounds like something I need to fix.

Also, what is the green diamond on this transfer plot above where I meet up with Eve?

http://i.imgur.com/qLZ8q0t.png

Green diamond is the transfer orbit periapsis.

Also, also - Arrowstar do you have any general advice to share about planning with the FMS? Obviously there's no real defined process like there would be with a point-to-point mission, but surely there are some things one should not do when setting up runs with the FMS to avoid repeated "unfeasible results"? I'm just not sure what are some good parameters to define from the start and then how best to tweak things to work towards a result that is worth transferring over to MA to play around with. After 5 runs with "solution may not be feasible" I'm left wondering - well is this just impossible altogether regardless of the constraints I set or do I just need to change these constraints a little to make it work? And how?

I'm also keen on plotting gravity assist missions in the future as well, but my brain kinda melts a little bit when I consider how to go about doing it.

If you're constantly getting "infeasible" solutions, then your best bet is to widen up your launch window and your min/max transfer orbit durations for each leg of the journey. Those are the variables in play, so you can't find a solution, you need to widen the bounds on them a bit.

As for how to actually use FMS with MA, here's my process:

1. First, you need to get a solution with FMS. Once you have this, keep FMS around in the background so you can reference it.
   
2. Open up MA.
   
3. Set the epoch of your initial state to the departure date of your FMS solution, minus the rough time needed to leave the SoI of the departure body.
   
4. Set the orbit of the initial state to the initial orbit you used in FMS.
   
5. Insert a coast into MA to the first burn true anomaly specified in FMS.
   
6. Insert an impulsive burn into MA with the component magnitudes (prograde/normal/radial) shown in FMS.
   
7. Attempt to propagate to the periapse of the flyby body (this will fail as you likely won't hit the flyby body's SoI).
   
8. Here's where it gets tricky: you need to try to hit the same flyby orbit parameters in MA as shown in FMS. So if your flyby orbit has an SMA/ecc/inc/RAAN/arg of a/b/c/d/e, you need to try to get a flyby that also has those parameters (they can be off a bit due to the inexact nature of FMS, but they'll still be close). Use the optimizer to achieve this.
   
9. Insert a burn @ periapse if needed.
   
10. Insert another coast to the periapse of your next flyby body. Optimize to hit your next target.
    
11. Repeat 6-9 for each flyby that occurs.
    

That's the general process. Don't get me wrong, it's still going to be VERY hard to get more than one flyby to work, and even one is a challenge. But with some luck and persistence, you'll get it.

Does that answer the question? Anything else I can try to clarify?

[Drew Kerman]

Can you tell me how to reproduce this? Sounds like something I need to fix.

Load up MA, create a UT coast event, open the Time/Date helper window, click the X on the Edit State window to close it. Now Save/Cancel out the Time/Date helper window. Should screw up or close out KSPTOT

Does that answer the question? Anything else I can try to clarify?

Good for now!

New Stuff

is it possible to display the pop-up tooltip with extended orbital data for the Other Craft window?

I'm working on splitting a burn of 848.589m/s - I will be doing the burn on two separate engines since I have a stage to dump in between. If I use the maneuver splitter, I can enter in the precise amount of Dv I have remaining in my first stage, and then hit Save. But if I remove the automatic coast event I find the Pe of my end state has gone from my targeted 959km to 1138km. Furthermore when I open the Edit State window for the maneuvers they show the exact same values for all three burn fields, even though the Dv budget window says burn 1 is 542.596m/s and burn 2 is 305.993m/s

[Arrowstar]

Load up MA, create a UT coast event, open the Time/Date helper window, click the X on the Edit State window to close it. Now Save/Cancel out the Time/Date helper window. Should screw up or close out KSPTOT

Thanks, fixed.

is it possible to display the pop-up tooltip with extended orbital data for the Other Craft window?

You mean in the window where you enter the data? Or somewhere else?

I'm working on splitting a burn of 848.589m/s - I will be doing the burn on two separate engines since I have a stage to dump in between. If I use the maneuver splitter, I can enter in the precise amount of Dv I have remaining in my first stage, and then hit Save. But if I remove the automatic coast event I find the Pe of my end state has gone from my targeted 959km to 1138km. Furthermore when I open the Edit State window for the maneuvers they show the exact same values for all three burn fields, even though the Dv budget window says burn 1 is 542.596m/s and burn 2 is 305.993m/s

Can I see a MAT file? I suspect you're applying radial or normal delta-v and that's sliding the periapsis around.

[Drew Kerman]

You mean in the window where you enter the data? Or somewhere else?

The window where you enter the data. I realize now that it's not a great idea since the values can be changed on the fly right then and there. Maybe a right-click option on the spacecraft name in the list that opens up a State window like the kind you can get from mission events.

Can I see a MAT file? I suspect you're applying radial or normal delta-v and that's sliding the periapsis around.

I've already redone it and lost the original burn, but it was indeed applying radial and normal values to the maneuver. I figured the burn splitter would handle that though. What I did was just delete both the coast and second maneuver and just created and optimized my own second burn using the remaining Dv as a starting point. I'm still confused however as to why both maneuvers showed the exact same Dv values (which is why I had to delete the second one) - you may want to see if you can reproduce that

[bk2w]

That would be forum user "Bk2w", on his own initiative typically.

My apologies, but after a dead laptop, it's taken me quite a while to get it fixed and back to a state where KSP is even a possibility. I'm finally back, and looking at updated KSPTOT for OSX.

Sadly, I do not yet have access to 2014b, so while I can work on an OS X version of 1.1.9, I cannot do anything for 1.2.0. Hopefully that will change at some point, but it may be quite a while.

-bk2w

[GrihanOs]

Does that answer the question? Anything else I can try to clarify?

Hi,

Your explanations answered me too, thank you ,

I managed to plan a flyby mission, even though I might lack the piloting and engineering skills to execute it.

Small nitpick:

When the multi-flyby sequencer calculation is running, there seems to be no way to stop it:

The "Stop" button does nothing, and when closing the window with the X in the upper right the confirmation appears, but if I click "Yes", the calculation is just restarted.

Thank you again for your work!

[Drew Kerman]

hey Arrowstar I have an issue with the Mission Animator. Seems telling it to show other spacecraft kills it completely and I have to restart the application. Here's all I found in the log after exiting:

Error while evaluating TimerFcn for timer 'timer-1' 

CELL2MAT does not support cell arrays containing cell arrays or objects.

Again, works fine until I tell it to show other space craft. You can try to reproduce using this mission file. Just move the playback to Mun's SOI and then start playback, then check the option and it should seize up.

[Drew Kerman]

also, might you consider adding some multi-burn functionality to the Maneuver Execution Assistant? I usually do all these calculations in the game, but since the Kerbulator mod doesn't seem to be supported anymore it might crap out on me at some point (1.0 and beyond) so I'd like to begin using the MEA more often. Anyways, in particular I'm referring to my previously-described scenario where I'm executing a maneuver with two different engines and a large change of mass in-between when I dump the stage, currently I have to set up the MEA for two separate burns then add the burn times together to get my total burn time/split

Some more MEA usage notes:

- Is there a way to save the system level properties between runs of the program? I tend to use a lot of the same engines on my craft. Or maybe directly access any Thrusters from an open MA mission?

- If we paste orbital data from the clipboard, can that set the planet as well? (same for the RMS, I forget to do that all the time)

- Being able to import a maneuver from KSP to the MEA would be nice, as well as a means of copying over a maneuver from MA - currently I have to copy/paste out the values from the DV Maneuver state to Notepad so I can copy them back into MEA once I close the Edit State window. Then I have to open up the View State After Event window to get the true anomaly for the burn and enter it manually. It's just a lot of opportunity to screw things up

Edited February 10, 2015 by Gaiiden

[Arrowstar]

My apologies, but after a dead laptop, it's taken me quite a while to get it fixed and back to a state where KSP is even a possibility. I'm finally back, and looking at updated KSPTOT for OSX.

Sadly, I do not yet have access to 2014b, so while I can work on an OS X version of 1.1.9, I cannot do anything for 1.2.0. Hopefully that will change at some point, but it may be quite a while.

-bk2w

No worries. If you ever get a hold of R2014b and want to try building V1.2.0+, let me know.

Hi,

Your explanations answered me too, thank you ,

I managed to plan a flyby mission, even though I might lack the piloting and engineering skills to execute it.

Small nitpick:

When the multi-flyby sequencer calculation is running, there seems to be no way to stop it:

The "Stop" button does nothing, and when closing the window with the X in the upper right the confirmation appears, but if I click "Yes", the calculation is just restarted.

Thank you again for your work!

Yeah, the stop button is actually a MATLAB thing and I don't really know how to make it go away or cease it from rendering. I am aware it doesn't work, however. Thanks for bringing it up, I'll see if maybe I can't fix it eventually.

hey Arrowstar I have an issue with the Mission Animator. Seems telling it to show other spacecraft kills it completely and I have to restart the application. Here's all I found in the log after exiting:

Error while evaluating TimerFcn for timer 'timer-1' 

CELL2MAT does not support cell arrays containing cell arrays or objects.

Again, works fine until I tell it to show other space craft. You can try to reproduce using this mission file. Just move the playback to Mun's SOI and then start playback, then check the option and it should seize up.

Thanks, I'll take a look! I suspect it's an error born from the MATLAB version change.

also, might you consider adding some multi-burn functionality to the Maneuver Execution Assistant? I usually do all these calculations in the game, but since the Kerbulator mod doesn't seem to be supported anymore it might crap out on me at some point (1.0 and beyond) so I'd like to begin using the MEA more often. Anyways, in particular I'm referring to my previously-described scenario where I'm executing a maneuver with two different engines and a large change of mass in-between when I dump the stage, currently I have to set up the MEA for two separate burns then add the burn times together to get my total burn time/split

Some more MEA usage notes:

- Is there a way to save the system level properties between runs of the program? I tend to use a lot of the same engines on my craft. Or maybe directly access any Thrusters from an open MA mission?

- If we paste orbital data from the clipboard, can that set the planet as well? (same for the RMS, I forget to do that all the time)

- Being able to import a maneuver from KSP to the MEA would be nice, as well as a means of copying over a maneuver from MA - currently I have to copy/paste out the values from the DV Maneuver state to Notepad so I can copy them back into MEA once I close the Edit State window. Then I have to open up the View State After Event window to get the true anomaly for the burn and enter it manually. It's just a lot of opportunity to screw things up

So to be honest, I really don't like the MEA now that I have Mission Architect. I would say if you're looking to execute an impulsive maneuver in KSP, convert it to a finite duration maneuver by right clicking the mission script and selecting "convert impulsive maneuver" (when you have the correct maneuver highlighted). Then just execute that (if you upload a finite duration maneuver, the node is created at the start time). MA does everything MEA does but without the annoyances you pointed out above.