[WIN/MAC/LINUX] KSP Trajectory Optimization Tool v1.6.9 [New MATLAB Version!]

[futrtrubl]

Yeah it's definitely not stock. I'll make and test a stock (+mechjeb) save and if I can repeat the issue will send it.

Edward

[futrtrubl]

Haven't yet made a clean install to do the stock test but I did find an error while fooling around. In the rendezvous maneuver sequencer (RMS from now on), select orbiting about the Sun, for initial and final orbit select Kerbin and then add 180degs to the mean anomaly for the final orbit (in effect try to get to L3 incorrectly since the RMS is not the right way to get out of an SoI). Gives the error "Undefined function 'coder.internal.emlrtDumpStack' for input arguments of type 'struct'."

Screenshot here

Edward

[Arrowstar]

Yeah it's definitely not stock. I'll make and test a stock (+mechjeb) save and if I can repeat the issue will send it.

Edward

Great, thanks. I'll look into the other issue you reported, too. Thanks for the screenshot.

[Arrowstar]

Okay, so the issue in your screenshot:

It turns out you can't compute that rendevous because the initial and final orbits are more or less the same. This leads to an ambiguity that confuses the code and throws the error you saw (side note: the actual error you're seeing is because some function isn't being packed with compiled KSPTOT, but that's being thrown because the missing function is an error handling function...). In the next release of KSPTOT, I've added a check that will ensure this condition is caught and the user notified if so.

Thanks for the report!

I'm still looking into the first problem you reported, but I'm having trouble reproducing anything. Let me know when you have that save game ready. That will help a ton.

[Starstrider42]

I haven't had a chance to try out the new version yet, but does MA support (or can it be made to support, if it doesn't yet) loading asteroids from a save file and including them as destinations?

[Arrowstar]

Yeah it's definitely not stock. I'll make and test a stock (+mechjeb) save and if I can repeat the issue will send it.

Hey so I think I know what the issue with RMS is. The burn UT is something that I actually know in the code because the user provides epochs for the orbits and whatnot, but for some silly reason I've never displayed it in the GUI. So basically you might just be burning at the wrong UTs (no fault of your own, the application should be displaying that info and it's not). Anyway, I added the burn UTs to the GUI under the burn true anomaly information. The "fix" (showing burn UT) is displayed now and will be in the next release.

I haven't had a chance to try out the new version yet, but does MA support (or can it be made to support, if it doesn't yet) loading asteroids from a save file and including them as destinations?

Ah, well, funny you ask! So the answer is that in the released version of Mission Architect, the answer is no. There is no way to import asteroids into the application. In fact, right now, the best way to figure out how to get to an asteroid is to use the Rendezvous Maneuver Sequencer application.

However, in the development version, well, that's a different story. Two days ago I anticipated this very need (along with trying to dock with other ships, and RemoteTech users trying to come up with optimal comm constellations) and developed a concept in MA called "Other Spacecraft." These are really just points along an orbit that you can use for various tasks. For example, lets say I wanted to meet up with a vehicle in Low Kerbin Orbit. I can use Other Spacecraft to "add" this vehicle to my Mission Architect mission plan, and then use the optimizer to minimize the distance between my spacecraft and this other vehicle. Then, to verify I'm getting what I expect, I can use a completely new application within Mission Architect called Graphical Analysis (so called because it generates plots ) to produce the following:

Notice how the distance to the reference spacecraft drops? That's because I asked the optimizer to give me that solution. And yes, you could use this for asteroids, too! There's no difference between a spaceship and asteroid as far as Mission Architect is concerned.

So be patient, I'll probably get this out this weekend and you can try it and let me know what you think. Sound good?

[Starstrider42]

So be patient, I'll probably get this out this weekend and you can try it and let me know what you think. Sound good?

Not sure I understand the vertical lines on the plot, but yes, it does sound good.

Edited May 10, 2014 by Starstrider42

[Arrowstar]

So the plot is just underlining my point, which is two-fold: 1) MA supports modeling the orbits of other spacecraft and asteriods; and 2) MA supports coming up with orbits such that you can rendezvous with those other spacecraft. The plot just shows that the distance from the mission spacecraft to the "reference spacecraft" (AKA the target) is getting smaller as the mission progresses, and at the end of the mission, the mission spacecraft has met up with the target spacecraft.

(The vertical red lines represent DV maneuvers and the blue lines represent apoapsis crossings (numbered A1, A2, etc...). Green lines are for periapsis crossings, but I don't have any of those in this mission apparently.)

(EDIT: So it just occurred to me that my "be patient" line might have sounded more snippy than anything else. Didn't mean to come across like that, was just asking for patience/time to get the new features rolled out properly. )

Edited May 10, 2014 by Arrowstar

[Arrowstar]

Hi everyone!

This afternoon I'm happy to announce the release of KSP TOT v0.12.2! The change log for this release is provided below. This is primarily an update for Mission Architect (plus a few fixes elsewhere). The major changes to this version of KSP TOT are many-fold, but the two I'd like to highlight for everyone are the Graphical Analysis tool within MA and the ability to model ground stations and other spacecraft orbits.

The KSP TOT download link has been updated in the first post. Please take a look and let me know what you think! I cannot emphasis enough how excellent this update is and how many new features it brings and completes within Mission Architect. I'm quite proud of what I was able to do here and I hope you'll take the time to enjoy it as I have.

Please let me know if you have any questions.

Version 0.12.2

• Added new constraints to the MA Mission Optimizer
  
  • Orbital Period
    
  • Non-singular H1
    
  • Non-singular K1
    
  • Non-singular H2
    
  • Non-singular K2
    

  [*]Added Mission Architect Celestial Body Catalog that displays celestial body information

  [*]Added ability to plot state and state-derived quantities using MATLAB plotting capabilities

  • Classical (Keplerian) orbit elements
    
  • Vector (position/velocity) elements
    
  • Equinoctial elements
    
  • Spacecraft masses (propellants, dry, total)
    
  • Spacecraft eclipse state
    
  • Other spacecraft state and state-derived values
    
  • Generates CSV output on command
    
  • Shows orbital events such as DV maneuvers and apse crossings on the plots
    
  • Numerous graphical plotting options
    
    • Colors, grids, line width, sub-plots, etc
      

  [*]Initial implementation of body rotation code

  • MA now computes longitude and latitude
    
    • Optimizer constraint
      
    • Graphical analysis variable
      
    • State readout output
      

  [*]Added "Other Spacecraft" and "Ground Stations"

  • User can define a spacecraft in orbit or a ground station on a body surface
    
  • Can be used for optimization ("minimize distance to s/c")
    
  • Can be used for plotting in Graphical Analysis
    

  [*]Added burn UT to Rendezvous Maneuver Sequencer

  [*]Improved MA SoI transition robustness for elliptical orbits

  [*]Increased optimizer speed for missions with many events

  [*]Fixed bug with retrieving orbit from KSPTOTConnect

  [*]Fixed issue with Rendezvous Maneuver Sequencer when trying to compute manenuvers between orbits with similar geometry

  [*]Fixed issue that could cause script execution to slow with many events

  [*]Fixed issue with MA mission plan file sizes being >10 MB

  [*]Fixed issue that caused saving and opening of MA mission plans to be slow

Edited May 11, 2014 by Arrowstar

[futrtrubl]

Hey so I think I know what the issue with RMS is. The burn UT is something that I actually know in the code because the user provides epochs for the orbits and whatnot, but for some silly reason I've never displayed it in the GUI. So basically you might just be burning at the wrong UTs (no fault of your own, the application should be displaying that info and it's not). Anyway, I added the burn UTs to the GUI under the burn true anomaly information. The "fix" (showing burn UT) is displayed now and will be in the next release.

Though I do like that the UT is now shown in the RMS, that isn't really the issue as I am using KSPTOTConnect to upload the maneuver to the game and I'm not entering it in manually.

Unfortunately 2 HDDs decided to die on me and until I get them replaced no KSP testing for me.

Edward

[Arrowstar]

Though I do like that the UT is now shown in the RMS, that isn't really the issue as I am using KSPTOTConnect to upload the maneuver to the game and I'm not entering it in manually.

Unfortunately 2 HDDs decided to die on me and until I get them replaced no KSP testing for me.

Edward

Okay, so maybe the issue is that it's uploading a "time past periapse" number and not a UT number? What if you copy the UT number from the RMS into the maneuver upload box and see what happens?

[Arrowstar]

Hi everyone,

I'd like to quickly announce a patch for KSPTOT v0.12.2. There isn't enough here to call it 0.12.3, but it features a few bug fixes and improvements that I'd like to push out to the community. The change log is below:

• Corrected an error in Mission Architect wherein "Reverse Time Errors" could be encountered while in an elliptic orbit. (This was the main reason for the patch.)
  
• In Rendezvous Maneuver Sequencer, three updates:
  
  • The UT of the burns are now shown in the maneuver information windows;
    
  • When you right click on the windows and select the context menu to upload the maneuvers, the maneuvers are uploaded w.r.t. the UT of the burn, not the time past periapse. This was really throwing people off and people were wondering why the rendezvous maneuvers weren't actually getting people to where they needed to go...
    
  • Burn delta-v vectors are now shown in Prograde-Normal-Radial order. This standardizes the order across KSPTOT and also matches what PreciseNode uses. This change is actually effective across all KSPTOT maneuver planning applications, as well.
    

  [*]In Mission Architect, UTs in the state output fields are now shown in decimal format and not scientific format.

  [*]Input fields in Mission Architect events should store more decimal places now (there might be floating point roundoff, sorry!).

  [*]A new Mission Architect field: in the Script menu, you can now upload selected DV_Maneuver events directly to KSP if you have the KSPTOTConnect plugin running.

  [*]Fixed an error in the Mission Architect Graphical Analysis tool related to CSV output of data. Also fixed an error w.r.t. the calculation of periapsis and apoapsis crossing points.

  [*]A few other nit-picky things that occurred to me when I should have been concentrating on being at work or driving or in the shower or the like.

You can get this latest version of KSPTOT (still being called v0.12.2) here or by using the link in the first post. Questions or comments? Please post them here for me to see!

Thanks, everyone!

[TranceaddicT]

Arrowstar,

Just found this and I think what it does is amazing. Unfortunately, I seem to have a chair-keyboard interface error.

If I grab an 'Earliest Departure Time' directly from KSP and enter an 'Earliest Arrival Time' approx. 30d later, should I not get a plot that has a departure time after my original departure?

However, I'm getting Optimal Departure time which occurs BEFORE the date I have entered as 'Earliest Departure Time', and the graphic displays the exact same date/time as was entered.

Example:

http://www.screencourier.com/image/00xkeaevtncowlgs0hpfft3hew

[Arrowstar]

Arrowstar,

Just found this and I think what it does is amazing. Unfortunately, I seem to have a chair-keyboard interface error.

If I grab an 'Earliest Departure Time' directly from KSP and enter an 'Earliest Arrival Time' approx. 30d later, should I not get a plot that has a departure time after my original departure?

However, I'm getting Optimal Departure time which occurs BEFORE the date I have entered as 'Earliest Departure Time', and the graphic displays the exact same date/time as was entered.

Example:

http://www.screencourier.com/image/00xkeaevtncowlgs0hpfft3hew

http://www.screencourier.com/image/00xkeaevtncowlgs0hpfft3hew

Believe it or not, you actually found a bug. Incredible, too, since this one has been around for over a year. Anyway, you're right, the bounds of the optimum solution should be constrained by the minimums you set. I've updated the code to account for that.

Please go ahead and re-download the KSPTOT package from the same place you got it before (first post of this thread works great). Let me know if that resolves the issue for you or not. And if I can help with any other questions, please ask! KSPTOT can have a bit of a learning curve, just like KSP, so don't be shy.

[krmcginnis]

Arrowstar,

I seem to be having an issue calculating the initial departure and arrival times for all planets. When I compute the porkchop plot I get something like the image shown below. The data point on the plot makes sense but the text output at the bottom has the arrival time BEFORE the departure time. When I click "Generate Departure Burn" the departure time and arrival times are auto-populated from the incorrect text output.

http://imgur.com/IyCOnnH

Edited May 16, 2014 by krmcginnis
Fixed image link.

[Arrowstar]

Yup, it's a bug. I've corrected the issue and will upload a build this afternoon when I get home from work. Thanks for the report!

[Drew Kerman]

at least for me, on Win7 x64, it required a restart to get KSPTOT to recognize that MCR was installed. MCR didn't tell me I needed to restart. Might want to add that as a note to the install instructions

[PlonioFludrasco]

Gonna upload that!!

[Starstrider42]

at least for me, on Win7 x64, it required a restart to get KSPTOT to recognize that MCR was installed. MCR didn't tell me I needed to restart. Might want to add that as a note to the install instructions

For the record, I haven't had any MCR problems, and I also use 64-bit Windows 7.

[diomedea]

at least for me, on Win7 x64, it required a restart to get KSPTOT to recognize that MCR was installed. MCR didn't tell me I needed to restart. Might want to add that as a note to the install instructions

Have you tried just relaunching KSP TOT? On my machine (Win8 64bit) KSP TOT always fails to recognize MCR the first time it is launched since reboot; but it always opens right the second and subsequent tries (I have no idea why, just letting know in case it may come useful).

[Drew Kerman]

Have you tried just relaunching KSP TOT? On my machine (Win8 64bit) KSP TOT always fails to recognize MCR the first time it is launched since reboot; but it always opens right the second and subsequent tries (I have no idea why, just letting know in case it may come useful).

I tried launching multiple times after installing MCR and kept getting the error that the MCR DLL was not found. Once I rebooted the computer (had to install some Windows updates anyways) it launched on the first try

[Arrowstar]

Gonna upload that!!

Err, sorry? What do you mean?

Have you tried just relaunching KSP TOT? On my machine (Win8 64bit) KSP TOT always fails to recognize MCR the first time it is launched since reboot; but it always opens right the second and subsequent tries (I have no idea why, just letting know in case it may come useful).

Yup, no idea either. It may just be an issue with the MCR...

I tried launching multiple times after installing MCR and kept getting the error that the MCR DLL was not found. Once I rebooted the computer (had to install some Windows updates anyways) it launched on the first try

So, restarting your PC after installing MCR is more or less a standard step. However, it's also not unusual to have to restart a PC after installing software of any kind, so I rather thought this was understood. Or do you mean you have to restart again, after the restart to install MCR?

[Drew Kerman]

So, restarting your PC after installing MCR is more or less a standard step. However, it's also not unusual to have to restart a PC after installing software of any kind, so I rather thought this was understood. Or do you mean you have to restart again, after the restart to install MCR?

I generally don't like to bother restarting after installing something just because I installed something. If it doesn't work, then of course the first thing I try is restarting the computer before doing something like trying to uninstall and re-install - or post for help on a forum. But that's me. I'm just saying this is probably something you should add to the FAQ item about the program not starting.

[Arrowstar]

Hi all,

Just a note that I've uploaded a new executable that fixes a few bugs people have found over the past 24 hours:

• Fixed two bugs with the porkchop plot application regarding bounds;
  
• Minmus orbital information updated such that argPeri = 0 and mean = ~89.5662 degrees in order to (hopefully) get the computed position of Minmus to match KSP;
  
• MA Constraints on RAAN, ArgPeri, True Anomaly, and longitude now are valid from -180 to 360 degrees (as opposed to 0 to 360 degrees);
  
• Updated the way MA coast optimization values are validated;
  
• Fixed an issue when saving an MA script with no constraints in it;
  
• Fixed an issue in MA graphical analysis when plotting w.r.t. mission elapsed time; and
  
• Fixed an issue where latitudes are computed incorrectly.
  

Grab the new KSPTOT in the OP's download/install section. Please let me know if you have any questions! Thanks!

[Arrowstar]

I generally don't like to bother restarting after installing something just because I installed something. If it doesn't work, then of course the first thing I try is restarting the computer before doing something like trying to uninstall and re-install - or post for help on a forum. But that's me. I'm just saying this is probably something you should add to the FAQ item about the program not starting.

No worries. I've added a note to the download/install instructions in the OP. It's in bold text, should be hard to miss. ;-) Thanks for the tip.