[WIN] Flyby Finder V0.86 [KSP1.3]

[PLAD]

6 hours ago, smjjames said:

The first attempt I tried to go by the graph and go by a date which looked good, but it ended up not working very well since I was coming in from the wrong place (on the outbound path rather than inbound from the suns periapsis), but in using the SlingShotter thing to look at planetary positions, I THINK there might be a chance to alter the inclination at apoapsis to meet Moho. Don't know if I have the fuel for it though.

Edit: I found a solution to get to Moho. Now I'm trying to get a Jool and then Eeloo flyby (Voyager 2). I had the app working, but when I was trying to narrow it down to the closest launch window, it stopped showing that.

Flying by Eve to get to Moho is tricky so you are doing well in pulling that off. I flew a K-E-M just now to refresh my memory (Kerbin Y1 D6.0 to Eve Y1 D191.2 to Moho Y1 D317.5) and I had to get the Eve flyby timing to within +/- 1 hour of the given time for it to work. And then there were three possible Moho encounters depending on the flyby altitude and I couldn't get the encounters at all while still in Kerbin orbit since it required more accuracy than the +/-.01m/s that Mechjeb allows. But by getting close and then fine tuning just before the Eve flyby it worked. (LKO to LMO for 4318m/s.) With the little bitty SOI's of Moho and Dres I have to trust that the planet will be there when FF says it will. You gave me a good idea though, I'm going to try Slingshotter so I can see where the planet will be when my ship will pass close to its path, even when I don't pass through the AN or DN at the orbit intersect (since KSP doesn't show the closest approach in that circumstance). Good luck with K-J-Eeloo. From my experiences you can be a couple weeks off of the Jool flyby time and still make it work. Best to leave Kerbin within a few hours of the right time though.

    Zooming in sometimes causes a solution to disappear. Usually it is because a middle encounter goes outside its reduced search period.

[smjjames]

Yeah, anyway, what I had done was that I decided to just match planes with Moho and wait for an encounter, which I had checked and wasn't all that far off.

And yeah, course corrections are going to be part of it anyhow.

[tomf]

Could the next version show details of the intermediate orbits - i.e. the AP and PE?

I have found a Kerbin-Eve-Kerbin-Jool plan with days 24-216-1084-2857

However as the flyby seems extremely variable on slight alterations to the plan caused by imperfect burns, and the second encounter is going to take a couple of orbits, I am having difficulty in getting my course corrections right. I think knowing what orbit I should be aiming for after the first flyby would be a great help.

[PLAD]

On 2/4/2016 at 3:20 PM, tomf said:

Could the next version show details of the intermediate orbits - i.e. the AP and PE?

I have found a Kerbin-Eve-Kerbin-Jool plan with days 24-216-1084-2857

However as the flyby seems extremely variable on slight alterations to the plan caused by imperfect burns, and the second encounter is going to take a couple of orbits, I am having difficulty in getting my course corrections right. I think knowing what orbit I should be aiming for after the first flyby would be a great help.

Hello, I understand the suggestion, for a while I experimented with using the inter-flyby orbit as a guide to setting up the next flyby, but I found it is not the best way to go. In the interest of keeping the detail box simple and small I therefore do not put details of the orbits in there. The most important thing in executing the flybys is the periapsis time at each planet in the chain, if you going to arrive at the right time and on the right side of the planet you should  be able to do a correction shortly before the flyby that will send you right to the next flyby. Below is a run through  of the plan you give above to show what I mean. Note the trick I use for each flyby, I put a second node on the path at just the time that the ship should be flying by the planet and then make adjustments to the first node so that that second node is almost right at periapsis. (I don't do it for the final arrival because I usually don't care if that is a little early or late, just that it happens.)

 You say the second encounter is going to take a couple of orbits, that doesn't sound right. Flyby Finder is somewhat restricted in what it finds, your ship should never go more than one orbit between encounters. The Eve flyby should kick you way out so that it takes 2 years to come back in and flyby Kerbin. See below.

One might point out that I don't really need the Vinf in and Vinf out numbers for each flyby, but I like to use them to check that a flight is going properly.  I have yet to find a time when the arrival times are right but the Vinf is wrong though, so I might get rid of them eventually.

[okder]

interface (for simple entering values from FlyBy Finder) for mechjeb  is here

 

[drtedastro]

Again, great mod.  Thanks.

What directory should the .exe be put in?  I am asking because the program is not detecting the OPM addon for extra planets.

 

Cheers.

[PLAD]

12 hours ago, drtedastro said:

Again, great mod.  Thanks.

What directory should the .exe be put in?  I am asking because the program is not detecting the OPM addon for extra planets.

 

Cheers.

It shouldn't matter where the FF .exe is put, the program doesn't read from or write to anything else. To 'turn on' the OPM planets, first have FF version 0.80 or higher running, then press the "clear" button. The button "Select Plantary System" should now be active, and you can press it and select the OPM system. Then hit "done", and now the OPM planets are selectable from the planet selection  dropdown lists.

Here's a screenshot of the button and selection window:

http://imgur.com/kEyOOVJ

 

[drtedastro]

Thanks.

Working great now.

Cheers.

[PLAD]

On 2/10/2016 at 11:08 AM, okder said:

interface (for simple entering values from FlyBy Finder) for mechjeb  is here

 

This looks handy! It sounds like you take the departure time and first flyby time as inputs, then use a Lambert algorithm to find the details of the transfer orbits (Two of them, the prograde and the retrograde solutions, FF weeds out the retrograde one but like you say it's easy to choose the right one.) Then you calculate the best time and inclination to launch from the KSC to get into that path, and it sounds like you perform the launch. This would eliminate the need for a test probe and my 'shift the node time back and forth until you get an encounter' trick.

I've downloaded it and will try it out. I'll put further comments in the thread you made for it.

[okder]

37 minutes ago, PLAD said:

FF weeds out the retrograde one but like you say it's easy to choose the right one

i for some reason did think that some times other solution can be useful and so there is checkbox for manual filling, i would probably think about elimination (based on minimum escape speed)

Quote

and it sounds like you perform the launch.

actually i do relay that to mechjeb code (with my own launch delay algo, mechjeb one is buggy), and have to warn that now inclined launch don't respect planet rotation speed, i.e. it goes to wrong inclination (could be easy manually corrected in upper layers of atmosphere up to 0.1 degree,while autopilot is active, if you reserve enough time in "LaunchOffset"), mechjeb devs said that bug is fixed in dev brunch, but i did not test it.

 

Experiments shows that prograde burn  in proposed orbit position works better than auto node creation by mechjeb code(with fixed arrival time), even if orbit slightly different than proposed one.

p.s. you could use inverse of that:

       s:= (tempy*tempz)/(2*tempz-tempy*transfers[boda,7]*transfers[boda,7]);    {SMA at first planet}

because that is infinity for parabola(and so your code gets instability if trajectory is very close to parabola), but inverse is correct one (i.e. 0 for parabola) for ellipse, parabola, hyperbola in your code.

 

Edited February 11, 2016 by okder

[PLAD]

You are right that that code could not handle an orbit close to a parabola. But that line only gets fed the results of a Lambert path from one planet to another, and in the stock system that means it will always be well above parabolic energy. For instance the minimum speed at infinity (true infinity, not the SOI) for a path from Kerbin to Eve is about 800 m/s. If I ever tried to use the code for going, say,  between Jool's moons it would crash hard. Of course in that case the fact that I used linked conics rather than patched conics would also not work. I optimized the program for a rather restricted set of orbits, but I think it does what it does do well. My greatest weakness right now is the start orbit definition, and I see you've beaten that with your superb addon. I need to at least give the LAN of the optimal start orbit. Next version.

[PLAD]

A note for Flyby Finder users, I've tried okder's addon, and it is superb for knowing when to launch to get into the best start orbit, and for setting up the node to depart from your start planet and get to the first encounter planet at the right time. It even corrects for errors in getting into the start orbit! I put a quick review on his thread. It's not perfect yet, but then neither is FF!

[okder]

i just saying that you could use instead of s 1/s (i.e. calculate 1/s firstime, look my code) and that would fix problem with parabola.

P.S. @PLAD, you forgot to press Choose button in review, that's why you got 180 degree inclination for duna.

 

[AlexTheNotSoGreat]

I do find this tool to be quite interesting, there are a couple things I feel would make this more useful...

   - Rather than a .exe file, make it a mod so it can be used in game, rather than having to leave and reenter it over and over again

   - RSS compatible, where flybys are a must have

I'm sure you have a reason on why you haven't done these, but go ahead and take your time!

[a_schack]

10 hours ago, AlexTheNotSoGreat said:

I do find this tool to be quite interesting, there are a couple things I feel would make this more useful...

   - Rather than a .exe file, make it a mod so it can be used in game, rather than having to leave and reenter it over and over again

   - RSS compatible, where flybys are a must have

I'm sure you have a reason on why you haven't done these, but go ahead and take your time!

PLAD did make an RSS version of it. It's in the bottom of the first post, but here's a direct link: 

And okder made a MechJeb plugin, though it requires the values from Flyby Finder. 

 

[PLAD]

What a_schack said. I might add that a reason that I don't want to put it in game is that it takes too big a window to show all its info, especially when I'm searching through the table. I tend to spend a long time searching for the best path and when I finally settle on one I copy it as text from the detail box, which is easy to refer to later.

   The other big reason is that I don't have the programming chops to put it in game.

[BevoLJ]

Small quality of life request.  Might it be possible to get a decrement button next to the increment button? =)

[PLAD]

6 hours ago, BevoLJ said:

Small quality of life request.  Might it be possible to get a decrement button next to the increment button? =)

That's a good idea, and not hard to make. Hmmm... there are a couple of other small tweaks I've been thinking about, I'll push out a small upgrade in week or so.

[PLAD]

I've just released version 0.83. It adds a "subtract" button to the 'Change all search dates by' field to go along with the "add" (was "increment") button. I also added a new field to the detail box which gives the start boost for a given path if you were to start in an equatorial orbit instead of an inclined orbit. There was a discussion about this in another thread some time ago, now you can compare the two ways to see if it's worth the extra trouble of getting into an inclined orbit. This release also fixes a bug in the day converter box, it used to give a value for the UT days that was one too low. The outputs in the table, graph, and detail box were correct so this wouldn't have affected any flight calculations.

Edited March 31, 2016 by PLAD

[BevoLJ]

Awesome update!  Cheers.  =)

[Stone Blue]

@PLAD Just a note to say that odker's MechJeb Addon just got updated for MJ 2.5.7, and is now 1.1.x compatible.

[PLAD]

26 minutes ago, Stone Blue said:

@PLAD Just a note to say that odker's MechJeb Addon just got updated for MJ 2.5.7, and is now 1.1.x compatible.

Thanks for the note, I find his addon extremely useful for getting the most efficient launches easily and was looking forward to this.

[Stone Blue]

@PLAD, well, to tell you the truth, I've known of your mod, and Transfer Window Planner for QUITE some time, but I've been daunted by the interfaces, so i havent tried them... But I saw odker's MJ add-on, and I think I'll give yours a try now... Figured I'd post odker has updated, so maybe other people who also have been on the fence about your mod may want to give it a try now, too... Plus its been a couple months since your thread has been "bumped"...

Edited June 6, 2016 by Stone Blue

[subyng]

I'm trying to find a fast Kerbin-Eve-Duna-Kerbin route. In your examples, you have a Kerbin-Eve-Duna-Kerbin route that can take as little as 138 days, but FF is unable to find that solution when entering the same, or even wider search parameters.

 

 

[PLAD]

1 hour ago, subyng said:

I'm trying to find a fast Kerbin-Eve-Duna-Kerbin route. In your examples, you have a Kerbin-Eve-Duna-Kerbin route that can take as little as 138 days, but FF is unable to find that solution when entering the same, or even wider search parameters.

 

 

The numbers used in the primer are in "Earth Time" (24/365). So the earliest search date is day 160 Earth time, which is day 637 in "Kerbal time" (6h/426d). Versions of FF after 0.60 start by default in Kerbal time, but earlier versions only did Earth time. (This happened because when I first wrote FF, KSP only did Earth time.) In any case the quickest way to translate them is to enter the numbers in FF when it is in Earth time mode just as they are in the example, and then click the little radio button to change FF to Kerbal time  and proceed with the search. You will find the same solutions but now the dates and time of flights will be in Kerbin days, so that 138 (Earth) day flight will now take about 552 (Kerbal) days.

   Some parts of the primer still show screenshots from version 0.60, though I did put in a note that all screenshots up to the 'new features' section use Earth time. These two time systems have caught me more than a few times.