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

RadarManFromTheMoon · September 11, 2014

today i managed to plan a flyby at eve and duna using a single burn at kerbin and only a little correction burn at eve. Pretty cool! I'm one of those guys who tend to plan their missions ahead excessively before building the rocket according to that. It's not the kerbal way. But it works for me. So what I whould really love to see is some kind of mission briefing that concludes the burns and course times (for life support) in a mission and helps designing a proper vessel. Something like this:

---------------------------------
mission start
---------------------------------
Burn 1
total delta-v: 100 m/s
---------------------------------
Course to pe
course time 5m23s (xxs)
---------------------------------
Mission end
---------------------------------
total delta-v: xxx
total mission time: xxx

I know I can get most of that info by inspecting the items in the mission architect. But thats tedious.

I also tried the MCC Real Time System and I can connect to the KSPConnect plugin and download orbits from it. But as once as I try to upload a maneuver I get an error and it refuses to let me download orbits anymore aswell.

*** Generating Maneuver Node... ***
Connection to KSPTOT Connect failed: Write to KSPTOT Connect failed: Unsuccessful open: Connection
refused: connect

Edited September 11, 2014 by RadarManFromTheMoon

Arrowstar · September 11, 2014

I think the Graphical Analysis is missing an option to show distance traveled. I like how NASA and them can say how far their probes have traveled, but currently there's no tracking of that in KSP, at least not through multiple switches between craft (Flight Results window distance traveled stat resets each load).

I implemented this tonight in about 10 minutes. It'll be in v1.1.7.

today i managed to plan a flyby at eve and duna using a single burn at kerbin and only a little correction burn at eve. Pretty cool! I'm one of those guys who tend to plan their missions ahead excessively before building the rocket according to that. It's not the kerbal way. But it works for me. So what I whould really love to see is some kind of mission briefing that concludes the burns and course times (for life support) in a mission and helps designing a proper vessel. Something like this:
I know I can get most of that info by inspecting the items in the mission architect. But thats tedious.

It's not a bad idea. I'll think about it.

I also tried the MCC Real Time System and I can connect to the KSPConnect plugin and download orbits from it. But as once as I try to upload a maneuver I get an error and it refuses to let me download orbits anymore aswell.

It's probably your firewall or something else blocking the connection. Are you connecting KSP TOT to the same computer you're playing KSP on? Do KSP TOT and KSP both have firewall exceptions?

So, quick update about what I'm planning for v1.1.7. In addition to all the other features that I've mentioned either already implementing or that are under consideration, I believe the next version of KSP TOT will be moving to a 64-bit architecture (it's a 32-bit application now). Coinciding with this change, I'm going to consider upgrading from MATLAB 2013a to MATLAB 2014b if I think it's worth it. Two years is a long time in MATLAB-land and there are lots of new features that I may be able to take advantageous of. A look at the R2014b change log, whenever it is published, will be what decides it for me.

This change wouldn't really impact anyone except for the need to re-download the new 2014b MCR. Sorry, I know, this is a big download, but if I find it's worth going to the new MATLAB version, it'll be worth the download on your end. I promise.

Let me know if you have any more questions (or feature requests!).

Drew Kerman · September 11, 2014

Nice, I was indeed thinking it would be a simple thing.

Is there a way to copy data tips from the graphical analysis charts? I see an export option but don't understand what it does - tried it and didn't see anything as a result. Would be nice to have the option to copy just the X or the Y value (maybe as selections in the right-click menu)

Arrowstar · September 11, 2014

Nice, I was indeed thinking it would be a simple thing.
Is there a way to copy data tips from the graphical analysis charts? I see an export option but don't understand what it does - tried it and didn't see anything as a result. Would be nice to have the option to copy just the X or the Y value (maybe as selections in the right-click menu)

Unfortunately the data tips are a MATLAB feature that I borrow but don't have any control over. The export function actually spits out the variable to the internal MATLAB workspace, so while it would in theory be accessible to KSP TOT, it's not directly accessible to you the user. If you're looking to get data out, I would recommend checking the "Generate Tabular Text Output" box before you create the plots. This will create a CSV file with all the plot data in it and you can go wild analyzing it in Excel, Open Office, or what have you. Your data tip data points will be in the CSV file.

RadarManFromTheMoon · September 11, 2014

I set the rules in the firewall. It even gets a ceonnection but uploading the maneuver nodes messes something up. Even if I deactivate the firewall. This is the output:

*** Checking Connection Status ***
Connection to KSPTOT Connect succeeded!
*** Generating Maneuver Node... ***
Connection to KSPTOT Connect failed: Write to KSPTOT Connect failed: Unsuccessful open: Connection
refused: connect

edit:

I deactivated the firewall on both PCs.

Which ports do you use?

Edited September 11, 2014 by RadarManFromTheMoon

diomedea · September 11, 2014

Let me know if you have any more questions (or feature requests!).

A rather old feature request, but more actual than ever: possibility to select the times displayed not only in Earth units (365 d/y, 24 h/d) but in Kerbin units too (426 d/y, 6 h/d). Just for convenience, as that avoids to convert from/to UT everytime in KSP.

Drew Kerman · September 11, 2014

I would recommend checking the "Generate Tabular Text Output" box before you create the plots. This will create a CSV file with all the plot data in it and you can go wild analyzing it in Excel, Open Office, or what have you. Your data tip data points will be in the CSV file.

Excellent. Only problem though is it just tabulates the data vertices. So would plotting this in Excel or similar program give me the proper data inbetween? Or does the graphical analysis need to be able to output a finer granularity? Like for example, I can zoom in on the graph in MA and unsnap the cursor to get very exact values if needed

Also, is there a reason mission elapsed time doesn't actually start at 0?

Edited September 11, 2014 by Gaiiden

Drew Kerman · September 11, 2014

in other news, I had an asteroid on final approach to Kerbin (class B, phew!) so I went to see if I could use MA to tell me where it would land, when and how fast it would be moving at the time. It was all pleasingly accurate

Arrowstar · September 11, 2014

I set the rules in the firewall. It even gets a ceonnection but uploading the maneuver nodes messes something up. Even if I deactivate the firewall. This is the output:
edit:
I deactivated the firewall on both PCs.
Which ports do you use?

I do believe it's on port 8282 for what you're trying to do. Let me know if it works. I'll investigate some tonight if not.

A rather old feature request, but more actual than ever: possibility to select the times displayed not only in Earth units (365 d/y, 24 h/d) but in Kerbin units too (426 d/y, 6 h/d). Just for convenience, as that avoids to convert from/to UT everytime in KSP.

As you've mentioned, I get this one a lot. The problem is that the math for converting in between the UT seconds and the date/time representation is hard-coded in a few places (never though the system of time would change, lol). Going through and injecting the KSP TOT user-defined options structure into all of those places would get real messy real fast. I'll look again, maybe I can come up with a simpler work around, but no promises here unfortunately.

Excellent. Only problem though is it just tabulates the data vertices. So would plotting this in Excel or similar program give me the proper data inbetween? Or does the graphical analysis need to be able to output a finer granularity? Like for example, I can zoom in on the graph in MA and unsnap the cursor to get very exact values if needed
Also, is there a reason mission elapsed time doesn't actually start at 0?
http://i.imgur.com/BzT90G8.png

The plots Graphical Analysis produce share the exact same data as in the CSV file, down to the arrays of doubles used in the code. So basically if you're getting it in the plot, you'll get it in the CSV file. Does that answer your question? I'm a bit unsure of what you were asking.

in other news, I had an asteroid on final approach to Kerbin (class B, phew!) so I went to see if I could use MA to tell me where it would land, when and how fast it would be moving at the time. It was all pleasingly accurate
http://i.imgur.com/8ftmroS.png

Nice! Do you just propagate to 0 km altitude above Kerbin and then look at the lat/long of the asteroid in the final state output tool tip text?

Drew Kerman · September 11, 2014

The plots Graphical Analysis produce share the exact same data as in the CSV file, down to the arrays of doubles used in the code. So basically if you're getting it in the plot, you'll get it in the CSV file. Does that answer your question? I'm a bit unsure of what you were asking.

Here's what I mean. I've zoomed in on the graph and plotted multiple data points:

However in the CSV file the first two values are (237.5370581, 83490.08515) and (475.0741161, 83442.14493) - these represent the values you get when the "Snap to Nearest Data Vertex" option is chosen for Selection Style. So I get how that is used to create a graph with Excel, just connect the dots and you have a nice graph:

But I'm less interested in the visual and more in the specific data. While I've gotten Excel to plot me the above graph, I haven't yet figured out how to tease out the finer-detailed data like I can do with the zoomed-in plot in MA when I turn off vertex lock.

Nice! Do you just propagate to 0 km altitude above Kerbin and then look at the lat/long of the asteroid in the final state output tool tip text?

Nope, I just inserted the coast to Periapse and then used the graphical analysis tool to get all the info I needed

Also, speaking of lat/long - the longitude actually seems borked, or I can't figure out how to read it. If I make a graph for longitude, it scales from 0 to 360 instead of -180 to 180. What's up with that?

Edited September 11, 2014 by Gaiiden

RadarManFromTheMoon · September 11, 2014

I think I know whats going on. Maybe I have forgotten to make that clear but I was trying to connect to KSP, which runs on my PC, from my notebook which is in the same subnet. I just tried it again over the localhost interface and everything worked fine. Looks like a firewall problem, but firewalls are disabled. So my conclusion was that there is probably something hard wired to localhost and that seems to be the case indeed.


function uploadManButton_Callback(hObject, eventdata, handles)
% hObject    handle to uploadManButton (see GCBO)
% eventdata  reserved - to be defined in a future version of MATLAB
% handles    structure with handles and user data (see GUIDATA)
    writeToOutput(handles.statusText,'*** Generating Maneuver Node... ***','append');

.
.   
.

    try 
        [B]tcpipClient = createTcpIpClient(8282, 'Client', 'localhost');[/B]
        writeDataToKSPTOTConnect(dataType, data, 'd', tcpipClient);
        failed = false;
    catch ME
        msg = ['Connection to KSPTOT Connect failed: ', ME.message];
        failed = true;
    end

    if(failed == false)
        msg = 'Maneuver Node created!';
    end

    writeToOutput(handles.statusText,msg,'append');
end

Edited September 11, 2014 by RadarManFromTheMoon

jeffchrisope · September 11, 2014

A rather old feature request, but more actual than ever: possibility to select the times displayed not only in Earth units (365 d/y, 24 h/d) but in Kerbin units too (426 d/y, 6 h/d). Just for convenience, as that avoids to convert from/to UT everytime in KSP.

+100 - please please please! In fact, I'm not sure we even need Earth units, do we? The Kerbin units are probably more important, at this point. Without fixing this issue, I'm back to entering seconds everywhere.

AppleDavidJeans · September 11, 2014

Do I have to buy MATLAB in order to run this software?

RadarManFromTheMoon · September 11, 2014

Do I have to buy MATLAB in order to run this software?

Nope, but you must install the Matlab runtime. Which is free. There is a link in the first post of this thread.

Drew Kerman · September 11, 2014

In fact, I'm not sure we even need Earth units, do we?

Yea, the RSS crowd will need them. Or anyone using them even with the regular Kerbol system

Arrowstar · September 11, 2014

But I'm less interested in the visual and more in the specific data. While I've gotten Excel to plot me the above graph, I haven't yet figured out how to tease out the finer-detailed data like I can do with the zoomed-in plot in MA when I turn off vertex lock.

MATLAB is just doing a linear interpolation on the data, so I suspect you could just do the same thing in Excel. Can you try that and let me know if it works?

Also, speaking of lat/long - the longitude actually seems borked, or I can't figure out how to read it. If I make a graph for longitude, it scales from 0 to 360 instead of -180 to 180. What's up with that?

The longitude given is East Longitude from the "prime meridian." Thus, the values shown are the number of degrees from the prime meridian, moving in an easterly directly. This is how we do it in the Real World of astrodynamics, so it's how I've implemented it here.

Oh hey, btw, I wasn't able to reproduce your "MET not starting at 0" issue. Can you send me the *.mat file with the mission you're seeing the problem in? I'd like to take a look.

I think I know whats going on. Maybe I have forgotten to make that clear but I was trying to connect to KSP, which runs on my PC, from my notebook which is in the same subnet. I just tried it again over the localhost interface and everything worked fine. Looks like a firewall problem, but firewalls are disabled. So my conclusion was that there is probably something hard wired to localhost and that seems to be the case indeed.

Thanks for tracking that down. I'll try putting together a debug build for you and see if that solves the problem. In the mean time, can you find the IPAddressWhitelist.txt in the KSPTOTConnect folder (in GameData, you may need to go down a few directories to see it)? Go ahead and add the IP address of the PC running KSPTOT there (if you're on the same network, it'll be the 192.* ... address I suspect). See if that helps any. It may not, but it's worth a shot.

+100 - please please please! In fact, I'm not sure we even need Earth units, do we? The Kerbin units are probably more important, at this point. Without fixing this issue, I'm back to entering seconds everywhere.

Alrighty, I've thought up a way to do it that doesn't involve rewriting half the program. It is, sadly, super hacky, but let me give it a try and see if it works consistently. If it does, it'll be in v1.1.7.

RadarManFromTheMoon · September 11, 2014

Thanks for tracking that down. I'll try putting together a debug build for you and see if that solves the problem. In the mean time, can you find the IPAddressWhitelist.txt in the KSPTOTConnect folder (in GameData, you may need to go down a few directories to see it)? Go ahead and add the IP address of the PC running KSPTOT there (if you're on the same network, it'll be the 192.* ... address I suspect). See if that helps any. It may not, but it's worth a shot.

already done that Without a whitelist entry KSPTOT establishes no connection at all.

The problem is that KSPTOT does not establish a connection when trying to upload a Maneuver node. I've browsed a bit through the code and as far as I see, the problem is that the upload function tries to connect to localhost via "createTcpIpClient(8282, 'Client', 'localhost')". when there is no host parameter in this function call, "createTcpIpClient" will get the remote from a global variable "rHost". But if a remote host is defined it will set rHost to the new host (in this case localhost). from there on every attempt to connect will go to the loopback interface. Messing up everything and requires a fresh start of KSPTOT.

Arrowstar · September 11, 2014

Roger roger. I've removed the localhost input and have a debug build cooking. I'll shoot you a PM with a link when it's done in a few minutes. Can you check it out and get back to me tonight, do you think?

RadarManFromTheMoon · September 12, 2014

Roger roger. I've removed the localhost input and have a debug build cooking. I'll shoot you a PM with a link when it's done in a few minutes. Can you check it out and get back to me tonight, do you think?

With pleasure

Would be great to get that running. I'm a mod loving guy. I have no memory to spare. So running KSPTOT on the same machine is a nogo

Drew Kerman · September 12, 2014

MATLAB is just doing a linear interpolation on the data, so I suspect you could just do the same thing in Excel. Can you try that and let me know if it works?

I'll fiddle around some more and see. No idea how to use Excel for this, but don't mind learning - will just maybe take a few days to get around to it

The longitude given is East Longitude from the "prime meridian." Thus, the values shown are the number of degrees from the prime meridian, moving in an easterly directly. This is how we do it in the Real World of astrodynamics, so it's how I've implemented it here.

Well, ok that explains what it's doing but it still doesn't help me properly interpret the value. Is the prime meridian the same point on every body? Because it's not 0 degrees. I used the graphical analysis to see what time the ship hits 0 and snapped a pic of the screen at that time and saw that the ship was at 61.9879E longitude - over Mun anyways. If the program won't convert to the same format used in the game, it would at least help if the "prime meridian" as MA sees it were actually at 0 degrees longitude in the game

Oh hey, btw, I wasn't able to reproduce your "MET not starting at 0" issue. Can you send me the *.mat file with the mission you're seeing the problem in? I'd like to take a look.

Yea, uhm - I can't seem to repro it now either. I'll make sure to send you a file if it pops up again tho.

AppleDavidJeans · September 13, 2014

Alright so I'm trying to run the MATLAB installer on my Mac and it keeps crashing.

Here's the raw data from my Terminal app that was monitoring the attempted installs.

Last login: Sat Sep 13 12:36:53 on ttys000

/Users/david/Downloads/MCR_R2013a_maci64_installer-2/install ; exit;

Davids-MacBook-Pro:~ david$ /Users/david/Downloads/MCR_R2013a_maci64_installer-2/install ; exit;

Preparing installation files ...

Installing ...

Exception in thread "main" com.google.inject.ProvisionException: Guice provision errors:

1) Error in custom provider, java.lang.RuntimeException: java.lang.reflect.InvocationTargetException

at com.mathworks.wizard.WizardModule.provideDisplayProperties(WizardModule.java:68)

while locating com.mathworks.instutil.DisplayProperties

at com.mathworks.wizard.ui.components.ComponentsModule.providePaintStrategy(ComponentsModule.java:76)

while locating com.mathworks.wizard.ui.components.PaintStrategy

for parameter 4 at com.mathworks.wizard.ui.components.SwingComponentFactoryImpl.<init>(SwingComponentFactoryImpl.java:111)

while locating com.mathworks.wizard.ui.components.SwingComponentFactoryImpl

while locating com.mathworks.wizard.ui.components.SwingComponentFactory

for parameter 1 at com.mathworks.wizard.ui.WizardUIImpl.<init>(WizardUIImpl.java:62)

while locating com.mathworks.wizard.ui.WizardUIImpl

while locating com.mathworks.wizard.ui.WizardUI annotated with @com.google.inject.name.Named(value=BaseWizardUI)

at com.mathworks.wizard.ui.UIModule.provideWizardUI(UIModule.java:50)

while locating com.mathworks.wizard.ui.WizardUI

for parameter 0 at com.mathworks.wizard.ExceptionHandlerImpl.<init>(ExceptionHandlerImpl.java:22)

while locating com.mathworks.wizard.ExceptionHandlerImpl

while locating com.mathworks.wizard.ExceptionHandler

1 error

at com.google.inject.InjectorImpl$4.get(InjectorImpl.java:767)

at com.google.inject.InjectorImpl.getInstance(InjectorImpl.java:793)

at com.mathworks.wizard.WizardLauncher.startWizard(WizardLauncher.java:156)

at com.mathworks.wizard.WizardLauncher.start(WizardLauncher.java:71)

at com.mathworks.wizard.AbstractLauncher.launch(AbstractLauncher.java:27)

at com.mathworks.wizard.AbstractLauncher.launchStandalone(AbstractLauncher.java:18)

at com.mathworks.mcrinstaller.MCRInstallLauncher.main(MCRInstallLauncher.java:28)

Caused by: java.lang.RuntimeException: java.lang.reflect.InvocationTargetException

at com.google.inject.internal.ProviderMethod.get(ProviderMethod.java:106)

at com.google.inject.InternalFactoryToProviderAdapter.get(InternalFactoryToProviderAdapter.java:48)

at com.google.inject.InjectorImpl$4$1.call(InjectorImpl.java:758)

at com.google.inject.InjectorImpl.callInContext(InjectorImpl.java:811)

at com.google.inject.InjectorImpl$4.get(InjectorImpl.java:754)

at com.google.inject.spi.ProviderLookup$1.get(ProviderLookup.java:89)

at com.google.inject.internal.ProviderMethod.get(ProviderMethod.java:95)

at com.google.inject.InternalFactoryToProviderAdapter.get(InternalFactoryToProviderAdapter.java:48)

at com.google.inject.SingleParameterInjector.inject(SingleParameterInjector.java:42)

at com.google.inject.SingleParameterInjector.getAll(SingleParameterInjector.java:66)

at com.google.inject.ConstructorInjector.construct(ConstructorInjector.java:84)

at com.google.inject.ConstructorBindingImpl$Factory.get(ConstructorBindingImpl.java:111)

at com.google.inject.FactoryProxy.get(FactoryProxy.java:56)

at com.google.inject.SingleParameterInjector.inject(SingleParameterInjector.java:42)

at com.google.inject.SingleParameterInjector.getAll(SingleParameterInjector.java:66)

at com.google.inject.ConstructorInjector.construct(ConstructorInjector.java:84)

at com.google.inject.ConstructorBindingImpl$Factory.get(ConstructorBindingImpl.java:111)

at com.google.inject.FactoryProxy.get(FactoryProxy.java:56)

at com.google.inject.ProviderToInternalFactoryAdapter$1.call(ProviderToInternalFactoryAdapter.java:45)

at com.google.inject.InjectorImpl.callInContext(InjectorImpl.java:811)

at com.google.inject.ProviderToInternalFactoryAdapter.get(ProviderToInternalFactoryAdapter.java:42)

at com.google.inject.Scopes$1$1.get(Scopes.java:54)