Jump to content

Arrowstar

Members
 View Profile  See their activity

  • Posts

    2,562

  • Joined

  • Last visited

 Content Type 

Everything posted by Arrowstar

  1. Arrowstar
    All more cycles gives you is a higher probability of more fully exploring the trade space. So run as many as you have time for and leave it at that.
  2. Arrowstar
    100 is high. If I was doing it for real and not to troubleshoot, maybe only 10 or 15.
  3. Arrowstar
    So I ran 100+ cases last night of this MFMS set up and couldn't reproduce the error. Do you have a special bodies file in use or anything else not stock?
  4. Arrowstar
    Ah, I see! Yeah, in KSP the Sun's SoI is infinitely sized, so there's no bound where you could still be eccentric but "escape" if that makes sense.
  5. Arrowstar
    Okay, thanks for letting me know. If you can reproduce it, please provide the steps (as last time which was perfect) so I can investigate! In the example you provided me, it looks like you were missing a coast segment. Is that what you meant? As far as propagating hyperbolic Sun-centered orbits, you'll have to be careful with it. There is a maximum true anomaly for hyperbolic orbits, and it's possible that's what you're running into, or running over as the case may be. If you try to coast to a true anomaly greater than that maximum, the propagator will not be able to go to the target and MA will fail/skip that event. I was able to get it to propagate out to a true anomaly of 167 deg or so, but because the orbit starts so close to that point, you don't see much. And yes, I do get all notifications (replies to the thread and @ name notifications as well), so whatever you'd like to do is fine.
  6. Arrowstar
    Okay, this pre-release should handle these three requests. Please go ahead and give it an overview. I had to do a fair bit of work on the code for the options stuff, so if you find anything that doesn't work, please let me know. Thanks!
  7. Arrowstar
    Thanks! No, but it easily could. Would you like it to? It should... but I just tested it and it doesn't. Looks like a bug. I'll fix it! Thanks for the report. It doesn't but it could. Do you think it should? (That it, is that the intuitive response the user expects?) I'm at a loss to be honest. It definitely works for me, I made sure to test it before I uploaded the new executable. Could you maybe post a video or GIF somewhere that shows the behavior? I'd like to see what you're seeing.
  8. Arrowstar
    That sounds great. Just be sure to set aside a night or two for running all those sessions! And yes, if you see anything with MA and SoI transitions, please let me know. I need to actually get this pre-release out as an official release at some point.
  9. Arrowstar
    Okay, here's a new pre-release for KSPTOT v.1.5.8. This resolves the issue with the MFMS complex number bug described above, should prevent MFMS from stealing focus while running multiple runs, and implements a new, experimental algorithm to help resolve the Mission Architect SoI transition issue noted by @Drew Kerman. I would ask everyone who uses KSPTOT regularly to please give Mission Architect a try over the next few days and see how that works, particularly A) with finding SoI transitions and B) optimization performance. I did a bunch of testing myself yesterday and everything seems to work now, but the more eyes on this the better. Finally, a quick announcement looking towards the future. It's been 2 years since I've updated the MATLAB Compiler Runtime and I think that, with the next version (probably 1.6.0), I'll begin looking into moving to MATLAB R2017b. If anyone has any comments or concerns about this, please let me know ASAP. Thanks!
  10. Arrowstar
    I've resolved the issue with the MFMS code stealing focus, I believe. Hurrah! It was an easy matter of just side-stepping some internal MATLAB logic (more or less) and then keeping the figure window around to be re-written to instead of open/closed each time the GA stops and starts. The "go to function" code uses an algorithm to find the "zeros" (roots) of the function that correspond to the desired value. Here's a plot of the absolute value of your altitude over time (minus 70 km). Note the two points at which the plot is zero. Because of the two distinct zero points, it's a toss up as to which root the code will find. This is the problem, though it's not a bug, it's a property of the numerics. To mitigate this, use the "maximum propagation time" input in the coast dialog and set it to something before the second root, say 500 seconds. You should get the result you want. I'm still looking at the last item you shared. I'll hopefully have a new build out tonight.
  11. Arrowstar
    @Drew Kerman: Okay, I think I found the bug. There was an... issue (not sure I'd call it a bug directly) with some code in the Lambert solver that was producing complex numbers with a zero imaginary part (so, for example, 4 + 0i) for some reason that I don't understand. It only happened on rare occasions. These complex numbers propagated until they hit something that couldn't handle them and then the code exploded as we saw. I have hopefully fixed the issue. I'm going to look into the window focus issue and your mission file weirdness tomorrow hopefully. I might even have a new build to try out by then!
  12. Arrowstar
    I got an error like this. Is this what you see in your log? Error using atan2 Inputs must be real. Error in computeHyperOrbitFromMultiFlybyParams (line 46) hTheta = atan2(eccHat(3,:),thHat(3,:)); Error in multiFlybyObjFunc (line 139) [hSMAIn, hEccIn, hIncIn, hRAANIn, hArgIn, ~, ~, rp] = computeHyperOrbitFromMultiFlybyParams(smaIn, eIn, hHat, sHat, true); Error in multiFlybyExec>@(x)multiFlybyObjFunc(x,numRevsArr,bodiesInfo,celBodyData) (line 50) fitnessfcn = @(x) multiFlybyObjFunc(x, numRevsArr,bodiesInfo,celBodyData); Error in createAnonymousFcn>@(x)fcn(x,FcnArgs{:}) (line 11) fcn_handle = @(x) fcn(x,FcnArgs{:});
  13. Arrowstar
    Thanks for the steps, I'm trying them now. Someone above asked about the application taking focus while the Optimizer runs. I can look into it, but it may not be possible because it's internal MATLAB behavior. I'll see though.
  14. Arrowstar
    Screenshots with accompanied explanations would be the best bet. Please be as thorough as needed so I can exactly reproduce what you're doing. Thanks!
  15. Arrowstar
    You do need a central body called Sun, yes. As long as the body ID doesn't change, though, nothing should go out of sync. I'll take a look at the file this weekend to see if that change helps any.
  16. Arrowstar
    I definitely appreciate it! That part of the code is the most finicky part of MA, so any refinements are useful and good to know about!
  17. Arrowstar
    Well, you're certainly right! You've yet again managed to find another awful edge case in the SoI transition code. Congrats. Notice that when you first hit the Kerbin SoI you actually have an eccentric (not hyperbolic) orbit? This is because I have some code that actually searches for SoI transitions in a few kilometers into the SoI in order to help mitigate those funny cases where you bounce between SoIs because you're right on the bleeding edge. Normally it's too small to matter, but in this case, apparently it somehow does and the resulting state is eccentric and not hyperbolic. Anyway, the code wasn't set up to expect that (all downward transitions should be hyperbolic in theory), hence the funny issue. If you want to see where this is implemented, it's line 309 in the findSoITransitions() function. Anyway, I've created a new pre-release build that should resolve the issue. Give it a try and tell me what you think. Regarding the closing message, yes, I agree it's not intuitive. However, the main KSPTOT window is controlling the closing procedure, so there's not much I can do about it without lots of rework. I'll give it more thought though and see if another solution presents itself. @Three_Pounds - Did you still want a mean anomaly, true anomaly, eccentric/hyperbolic anomaly converter in the tools window as you originally requested? Or are you good now?
  18. Arrowstar
    Yes I can! Sorry life has been busy. I will get to you all in a day or two hopefully. In the meantime, no blasphemy please.
  19. Arrowstar
    You're welcome, glad to hear that everything works now!
  20. Arrowstar
    Okay, give this debug build a try and tell me if it resolves the issue: https://drive.google.com/file/d/0B83Py_-98MhoRkNQblNBVWRoNWM/view?usp=sharing
  21. Arrowstar
    Can I see the KSPTOT log file you have that shows the error? I opened up both of your MAT files and they loaded up fine to me.
  22. Arrowstar
    Ill try to take a look after work. What version of the software are you on?
  23. Arrowstar
    Okay, give this a try: KSPTOT v1.5.8 pre-release 6
  24. Arrowstar
    Oh... maybe not. I'll get it created tonight and post it here.
  25. Arrowstar
    I'm not totally sure what the PATH is either. I didn't write the Linux instructions, and I hope someone else can help here
×
×
  • Create New...