Advanced Question Help with trajectory planner!!!

downloaderfan

Member
Joined
Jan 12, 2013
Messages
104
Reaction score
0
Points
16
Greetings orbiter-forum community,
First, some backstory:
During my free time, I occasionally launch orbiter to find some cool 'launch windows' to perform slingshot maneuvers mostly to outer planets using assists from inner planets. I've posted one scenario titled 'The ultimate great grand tour' in the tutorials section of this forum and have many other scenarios which are just on my PC yet.
I very recently came across the Trajectory optimization tool after reading some posts, so I immediately downloaded the tool to find out whether it can simplify the process of finding launch windows for complex slingshot maneuvers. After reading the manual once, I went ahead to find some launch windows with some constrains that I've applied. I typically try to keep the fuel used as well as the TOF to a minimum, the fuel utilized is normally what it would take to get to venus, then I would sling around wherever else I wanted to go.

After coming up with a few trajectories in Trajectory optimization tool, I tried to plan the same in TransX but I was unable to do so due to a couple of reasons, that's why I'm writing this post in hopes of solving these issues.

So, the first trajectory I tried to find is a launch date for EVVEJ.

fan8uc.jpg


Zoomed in image

oibvyp.jpg


So you can see that, the transfer first from earth to venus looks fine, but after that, TOT doesn't display any orbit for venus to venus, then it again shows an orbit for venus to earth then jupiter. Is this some kind of limitation that TOT has? That it can't plan slingshots to the same planet again for gravity assists? While trying to plan this in orbiter, I was able to get from earth to venus with arrival at the specified date(as per TOT), also venus to venus at the specified date(although, I had to lose energy relative to the sun instead of gaining which is absurd), then got stuck when I tried to get to earth.

I would have shared the text data but it got lost due to a TOT crash so the image is all I have as of now, but if it's absolutely necessary, I don't mind planning the same trajectory again to post the text data. This just seems to be a general issue(getting to the same planet back) since some other trajectories I planned also had this issue of no orbit shown by TOT.

Problem 2:

Ok, so to avoid getting back to the same planet, this time I decided to plan EVMEJ.

So, again I was able to find a trajectory using TOT, but was not able to plan the same using transX in orbiter.

Here is an image of the TOT plan,

66xevr.png


Zoomed in

wtwvhk.jpg


Text data:

Starting optimization...
Optimization finished in 504.2767 seconds with message: Average change in trajectory
fitness function less than tolerance.
Results of the analysis are as follows:
The optimal departure from EARTH occurs at 6/12/2015 19:11:52 (C3=15.7723 km^2/s^2)
The optimal flyby of VENUS occurs at 9/14/2015 11:34:6 (deltaV=82.9022 km/s, pass
radius=26643.9676 km)
The optimal flyby of MARS occurs at 3/25/2016 2:55:36 (deltaV=154.3439 km/s, pass
radius=758.2502 km)
The optimal flyby of EARTH occurs at 1/28/2019 10:54:45 (deltaV=47.7631 km/s, pass
radius=7015.9559 km)
The optimal arrival at JUPITER BARYCENTER occurs at 7/3/2021 5:16:30 (arrival
velocity=6.1866 km/s)
The optimal trip duration is 2212.4199 days.

Again, while trying to plan this using TransX, I wasn't able to succeed. I was able to plan Earth to Venus very accurately.

abmmub.png


But again, getting to mars from venus was not possible using transX as depicted by TOT.

Here is the scenario where I have planned the trajectory from earth to venus, but getting to mars doesn't seem to be possible.

Code:
BEGIN_DESC

END_DESC

BEGIN_ENVIRONMENT
  System Sol
  Date MJD 57185.6090907301
END_ENVIRONMENT

BEGIN_FOCUS
  Ship GL-01
END_FOCUS

BEGIN_CAMERA
  TARGET GL-01
  MODE Cockpit
  FOV 60.00
END_CAMERA

BEGIN_MFD Left
  TYPE User
  MODE TransX
  Ship  GL-01
  FNumber 4
  Int 1
  Orbit True
  Vector  4377295.24512 1395005.12175 4413960.6041
  Vector  -254.847096252 -126.486384635 292.705407442
  Double  3.98600439969e+014
  Double  57185.6090904
  Handle Earth
  Handle NULL
  Handle NULL
Select Target
 0 Escape
Autoplan
0 0
Plan type
0 0
Plan
0 1
Plan
0 0
Plan
0 0
Select Minor
 0 None
Manoeuvre mode
0 0
Auto-Center™
0 0
Base Orbit
0 0
Prograde vel.
 1  0
Man. date
 1  57185.6090889
Outward vel.
 1  0
Ch. plane vel.
 1  0
Intercept with
0 0
Orbits to Icept
0 0
Graph projection
0 0
Scale to view
0 0
Advanced
0 0
Pe Distance
 1  7645212
Ej Orientation
 1  0
Equatorial view
0 0
Finvars
  Finish BaseFunction
  Int 2
  Orbit False
  Handle Sun
  Handle Earth
  Handle Venus
Select Target
 0 Venus
Autoplan
0 0
Plan type
0 2
Plan
0 0
Plan
0 0
Plan
0 1
Select Minor
 0 None
Manoeuvre mode
0 0
Auto-Center™
0 0
Base Orbit
0 1
Prograde vel.
 1  0
Man. date
 1  57185.6086678
Outward vel.
 1  0
Ch. plane vel.
 1  0
Intercept with
0 0
Orbits to Icept
0 0
Graph projection
0 0
Scale to view
0 0
Advanced
0 0
Prograde vel.
 1  -2951.38117033
Eject date
 1  57185.601335
Outward vel.
 1  -2147.21737164
Ch. plane vel.
 1  1520.28771854
Finvars
  Finish BaseFunction
  Int 4
  Orbit True
  Vector  6020403214.3 -173305917.758 -1302439487.84
  Vector  -5725.18527136 165.49087746 1191.97810921
  Double  3.2485863e+014
  Double  57267.0783594
  Handle Venus
  Handle NULL
  Handle NULL
Select Target
 0 Escape
Autoplan
0 0
Plan type
0 1
Plan
0 0
Plan
0 1
Plan
0 0
Select Minor
 0 None
Manoeuvre mode
0 0
Auto-Center™
0 0
Base Orbit
0 0
Prograde vel.
 1  0
Man. date
 1  57185.6086708
Outward vel.
 1  0
Ch. plane vel.
 1  0
Intercept with
0 0
Orbits to Icept
0 0
Graph projection
0 0
Scale to view
0 0
Advanced
0 0
View Orbit
0 0
Finvars
  Finish BaseFunction
  Int 3
  Orbit True
  Vector  106940198834 -5938074268.09 17048642654.4
  Vector  -10507.1821977 905.806012452 37703.0700697
  Double  1.32712764814e+020
  Double  57279.2023186
  Handle Sun
  Handle Venus
  Handle Mars
Select Target
 0 Mars
Autoplan
0 0
Plan type
0 2
Plan
0 0
Plan
0 0
Plan
0 2
Select Minor
 0 None
Manoeuvre mode
0 0
Auto-Center™
0 0
Base Orbit
0 0
Prograde vel.
 1  0
Man. date
 1  57185.6090889
Outward vel.
 1  0
Ch. plane vel.
 1  0
Intercept with
0 0
Orbits to Icept
0 0
Graph projection
0 0
Scale to view
0 0
Advanced
0 0
Velocity.
 1  0
Outward angle
 1  0.0383972435443
Inc. angle
 1  0
Inherit Vel.
0 0
Eject date
 1  57279.2023186
Finvars
  Finish BaseFunction
END_MFD

BEGIN_MFD Right
  TYPE User
  MODE TransX
END_MFD

BEGIN_SHIPS
ISS:ProjectAlpha_ISS
  STATUS Orbiting Earth
  RPOS 6481270.69 -1617261.61 -800244.29
  RVEL -1508.387 -7186.677 2320.934
  AROT 28.00 0.94 40.81
  AFCMODE 7
  IDS 0:588 100 1:586 100 2:584 100 3:582 100 4:580 100
  NAVFREQ 0 0
  XPDR 466
END
Mir:Mir
  STATUS Orbiting Earth
  RPOS 6329443.03 -125089.84 -2082000.96
  RVEL 2416.566 452.870 7338.343
  AROT 0.53 -44.47 89.92
  AFCMODE 7
  IDS 0:540 100 1:542 100 2:544 100
  XPDR 482
END
Luna-OB1:Wheel
  STATUS Orbiting Moon
  RPOS -44004.14 -2237543.73 -18.02
  RVEL 1479.820 -29.087 -0.002
  AROT 0.00 0.00 -167.96
  VROT -0.00 -0.00 10.00
  AFCMODE 7
  IDS 0:560 100 1:564 100
  XPDR 494
END
GL-01:DeltaGlider
  STATUS Landed Earth
  BASE Cape Canaveral:1
  POS -80.6758964 28.5227640
  HEADING 150.00
  AFCMODE 7
  PRPLEVEL 0:1.000000 1:1.000000
  NAVFREQ 402 94 0 0
  XPDR 0
  GEAR 1 1.0000
  AAP 0:0 0:0 0:0
END
SH-02:ShuttleA
  STATUS Landed Earth
  BASE Cape Canaveral:5
  POS -80.6745292 28.5197208
  HEADING 100.00
  AFCMODE 7
  PRPLEVEL 0:1.000000 1:1.000000
  NAVFREQ 0 0
  XPDR 0
  PODANGLE 0.0000 0.0000
  DOCKSTATE 0 0.0000
  AIRLOCK 0 0.0000
  GEAR 0 0.0000
  PAYLOAD MASS 0.0 0
END
END_SHIPS

BEGIN_ExtMFD
END

So I wanted to know if there's some mistake I'm doing as of now because I'm not able to find one from the manual. Thank you for your time.
 

boogabooga

Bug Crusher
Joined
Apr 16, 2011
Messages
2,999
Reaction score
0
Points
0
The problem is with your expectations.

A multiple inner planet slingshot sequence that makes sense occurs only rarely. The tough part of using TOT for this is that you have to know the sequence. TOT only knows the constraints that you put on it. It does not know to throw out a useless encounter or if some altogether better sequence exists.

Your first example is pretty much telling you to go orbit Venus for half a year or so and wait for another launch window to open, because you can't arrive there when you need to. In other word, it's junk. It is not the result that is absurd, it is your constraints.

In the second example, notice that each "deltaV" is huge. They should be near zero. You are not doing "slingshot" maneuvers, you are doing useless powered flybys of planets that are out of your way. Also, the radius of the Mars encounter is probably under the surface.

Practice running TOT on past known sequences at the proper time, such a Cassini, etc. Also, you can vet your trajectories with this:
http://trajbrowser.arc.nasa.gov/traj_browser.php

Other than that, you might consider enabling multi-revolution mode see if you need to make multiple orbits.

Edit:
For curiosity, I was able to find the Venus-Mars leg in TransX. Sure enough, it seems your problem was that you needed to turn off "inherit vel." and enter it in because you were not considering that you needed a (large) propulsive maneuver.

Code:
BEGIN_DESC
Contains the latest simulation state.
END_DESC

BEGIN_ENVIRONMENT
  System Sol
  Date MJD 57185.6397313285
END_ENVIRONMENT

BEGIN_FOCUS
  Ship GL-01
END_FOCUS

BEGIN_CAMERA
  TARGET GL-01
  MODE Cockpit
  FOV 60.00
END_CAMERA

BEGIN_MFD Left
  TYPE User
  MODE TransX
  Ship  GL-01
  FNumber 5
  Int 1
  Orbit True
  Vector  3625629.55471 1088178.76915 5124497.87448
  Vector  -311.284604813 -104.60584818 242.449612954
  Double  3.98600439969e+014
  Double  57185.6397292
  Handle Earth
  Handle NULL
  Handle NULL
Select Target
 0 Escape
Autoplan
0 0
Plan type
0 0
Plan
0 1
Plan
0 0
Plan
0 0
Select Minor
 0 None
Manoeuvre mode
0 0
Auto-Center™
0 0
Base Orbit
0 0
Prograde vel.
 1  0
Man. date
 1  57185.6397292
Outward vel.
 1  0
Ch. plane vel.
 1  0
Intercept with
0 0
Orbits to Icept
0 0
Graph projection
0 0
Scale to view
0 0
Advanced
0 0
Pe Distance
 1  7645212
Ej Orientation
 1  0
Equatorial view
0 0
Finvars
  Finish BaseFunction
  Int 2
  Orbit False
  Handle Sun
  Handle Earth
  Handle Venus
Select Target
 0 Venus
Autoplan
0 0
Plan type
0 2
Plan
0 0
Plan
0 0
Plan
0 1
Select Minor
 0 None
Manoeuvre mode
0 0
Auto-Center™
0 0
Base Orbit
0 1
Prograde vel.
 1  0
Man. date
 1  57185.6396891
Outward vel.
 1  0
Ch. plane vel.
 1  0
Intercept with
0 0
Orbits to Icept
0 0
Graph projection
0 0
Scale to view
0 0
Advanced
0 0
Prograde vel.
 1  -2951.38117033
Eject date
 1  57185.601335
Outward vel.
 1  -2147.21737164
Ch. plane vel.
 1  1520.28771854
Finvars
  Finish BaseFunction
  Int 4
  Orbit True
  Vector  6020438065.14 -173306460.853 -1302447101
  Vector  -5725.18423181 165.490665101 1191.97776498
  Double  3.2485863e+014
  Double  57267.078292
  Handle Venus
  Handle NULL
  Handle NULL
Select Target
 0 Escape
Autoplan
0 0
Plan type
0 1
Plan
0 0
Plan
0 1
Plan
0 0
Select Minor
 0 None
Manoeuvre mode
0 0
Auto-Center™
0 0
Base Orbit
0 0
Prograde vel.
 1  0
Man. date
 1  57185.6396984
Outward vel.
 1  0
Ch. plane vel.
 1  0
Intercept with
0 0
Orbits to Icept
0 0
Graph projection
0 0
Scale to view
0 0
Advanced
0 0
View Orbit
0 0
Finvars
  Finish BaseFunction
  Int 3
  Orbit True
  Vector  106940193197 -5938073192.78 17048659238.4
  Vector  -10507.200425 905.794681226 37703.0506284
  Double  1.32712764814e+020
  Double  57279.2023241
  Handle Sun
  Handle Venus
  Handle Mars
Select Target
 0 Mars
Autoplan
0 0
Plan type
0 2
Plan
0 0
Plan
0 0
Plan
0 2
Select Minor
 0 None
Manoeuvre mode
0 0
Auto-Center™
0 0
Base Orbit
0 0
Prograde vel.
 1  0
Man. date
 1  57185.6397292
Outward vel.
 1  0
Ch. plane vel.
 1  0
Intercept with
0 0
Orbits to Icept
0 0
Graph projection
0 0
Scale to view
0 0
Advanced
0 0
Velocity.
 1  10801.25
Outward angle
 1  -1.06408535704
Inc. angle
 1  -0.297323819394
Inherit Vel.
0 1
Eject date
 1  57279.202324
Finvars
  Finish BaseFunction
  Int 5
  Orbit True
  Vector  5785654100.11 935325665.716 -380648184.293
  Vector  -8188.38463088 -1324.13541162 539.291316575
  Double  4.28282991638e+013
  Double  57464.2982418
  Handle Mars
  Handle NULL
  Handle NULL
Select Target
 0 None
Autoplan
0 0
Plan type
0 1
Plan
0 0
Plan
0 2
Plan
0 0
Select Minor
 0 None
Manoeuvre mode
0 0
Auto-Center™
0 0
Base Orbit
0 0
Prograde vel.
 1  0
Man. date
 1  57185.6354005
Outward vel.
 1  0
Ch. plane vel.
 1  0
Intercept with
0 0
Orbits to Icept
0 0
Graph projection
0 0
Scale to view
0 0
Advanced
0 0
Draw Base
0 0
Finvars
  Finish BaseFunction
END_MFD

BEGIN_MFD Right
  TYPE User
  MODE TransX
END_MFD

BEGIN_SHIPS
ISS:ProjectAlpha_ISS
  STATUS Orbiting Earth
  RPOS -6591069.13 899519.18 1039540.08
  RVEL 655.313 7344.351 -2203.882
  AROT 26.34 -6.26 50.44
  AFCMODE 7
  IDS 0:588 100 1:586 100 2:584 100 3:582 100 4:580 100
  NAVFREQ 0 0
  XPDR 466
END
Mir:Mir
  STATUS Orbiting Earth
  RPOS -6182175.71 138551.00 2489960.45
  RVEL -2900.797 -447.226 -7158.537
  AROT 3.43 -46.76 88.26
  AFCMODE 7
  IDS 0:540 100 1:542 100 2:544 100
  XPDR 482
END
Luna-OB1:Wheel
  STATUS Orbiting Moon
  RPOS 2209236.46 357513.08 33.33
  RVEL -236.389 1461.135 0.041
  AROT 0.00 0.00 18.66
  VROT 0.00 0.00 10.00
  AFCMODE 7
  IDS 0:560 100 1:564 100
  XPDR 494
END
GL-01:DeltaGlider
  STATUS Landed Earth
  BASE Cape Canaveral:1
  POS -80.6758964 28.5227640
  HEADING 150.00
  AFCMODE 7
  PRPLEVEL 0:1.000000 1:1.000000
  NAVFREQ 402 94 0 0
  XPDR 0
  GEAR 1 1.0000
  AAP 0:0 0:0 0:0
END
SH-02:ShuttleA
  STATUS Landed Earth
  BASE Cape Canaveral:5
  POS -80.6745292 28.5197208
  HEADING 100.00
  AFCMODE 7
  PRPLEVEL 0:1.000000 1:1.000000
  NAVFREQ 0 0
  XPDR 0
  PODANGLE 0.0000 0.0000
  DOCKSTATE 0 0.0000
  AIRLOCK 0 0.0000
  GEAR 0 0.0000
  PAYLOAD MASS 0.0 0
END
END_SHIPS

BEGIN_ExtMFD
END
 
Last edited:

downloaderfan

Member
Joined
Jan 12, 2013
Messages
104
Reaction score
0
Points
16
Dear boogabooga, thanks for replying.

The problem is with your expectations.

A multiple inner planet slingshot sequence that makes sense occurs only rarely. The tough part of using TOT for this is that you have to know the sequence. TOT only knows the constraints that you put on it. It does not know to throw out a useless encounter or if some altogether better sequence exists.

Practice running TOT on past known sequences at the proper time, such a Cassini, etc.

I guess you've assumed here that I'm expecting a trajectory from TOT which I have not personally verified if possible or not with the constrains I've applied. As I've said, I already have many scenarios in which I've planned trips to outer planets using slings from venus using transX alone, so I was merely trying to get the same launch windows using TOT using the same sequences which I know very well, exist and work.

Your first example is pretty much telling you to go orbit Venus for half a year or so and wait for another launch window to open, because you can't arrive there when you need to. In other word, it's junk. It is not the result that is absurd, it is your constraints.

Ok, my constrains are absurd, but I absolutely have no idea what in my constrains could have caused something like that, where I get to another planet, wait for a while for it to spin around the sun, then leave that planet again instead of just slinging from it. I've made a plan similar to EVVEJ again, with more details this time.

2zfn1xd.jpg


My constrains applied: (A solution satisfying these constrains does exist, as I've successfully planned it using TransX alone)

352hzqs.jpg


Text data:

Starting optimization...
Optimization finished in 384.94 seconds with message: Average change in trajectory
fitness function less than tolerance.
Results of the analysis are as follows:
The optimal departure from EARTH occurs at 1/15/2017 22:27:12 (C3=9.3235 km^2/s^2)
The optimal flyby of VENUS occurs at 5/12/2017 5:29:45 (deltaV=15.5265 km/s, pass
radius=24757231393.664 km)
The optimal flyby of VENUS occurs at 11/12/2017 18:6:36 (deltaV=85.474 km/s, pass
radius=4615843444459.824 km)
The optimal flyby of EARTH occurs at 1/23/2020 7:11:21 (deltaV=0.070545 km/s, pass
radius=9643.722 km)
The optimal arrival at JUPITER BARYCENTER occurs at 9/23/2021 8:27:58 (arrival
velocity=10.0693 km/s)
The optimal trip duration is 1711.4172 days


Now obviously the above text data is absurd, with huge radius passes and high deltaV. Now ticking 'Constrain all flybys to 0 Delta-V' and 'Enable Multi-Rev Mode' maybe the solution to this absurdity but TOT always crashes when I try to obtain a solution with those 2 options ticked, that too after an hour of calculations. This has happened thrice now, I'm pretty sure it'll happen more times, not that anything can be done about that I guess.

145so7.jpg


In the second example, notice that each "deltaV" is huge. They should be near zero. You are not doing "slingshot" maneuvers, you are doing useless powered flybys of planets that are out of your way. Also, the radius of the Mars encounter is probably under the surface.

Edit:
For curiosity, I was able to find the Venus-Mars leg in TransX. Sure enough, it seems your problem was that you needed to turn off "inherit vel." and enter it in because you were not considering that you needed a (large) propulsive maneuver.

About the deltaV being huge, I've already stated that 'Constrain all flybys to 0 Delta-V' might solve that problem, but I still don't see a constrain that would prevent 'under the surface' slingshot solutions, which would again mean a powered slingshot over the surface. I always avoid powered slingshots to keep fuel utilization to a minimum, normally while planning using transX, if I see that a non powered slingshot trajectory doesn't take me to my destination, I start over again from a different date, that's the reason I didn't consider the powered slingshot option by turning off "inherit vel". Going from EVMEJ doesn't always require a powered slingshot, as seen from this flytandem scenario, and this is the solution I was expecting from TOT, a completely non powered over the surface slingshot.
 
Last edited:

boogabooga

Bug Crusher
Joined
Apr 16, 2011
Messages
2,999
Reaction score
0
Points
0
Okay. Actually, you are right and I found a reasonable solution to your EVVEJ problem:

Code:
      Trajectory Optimization Tool v2
                                            Written by Adam Harden ("Arrowstar"), (C) 2011
#####################################################################                     
WARNING: Number of bodies in Flight Plan altered: deleting stored Flight Constraints.     
Starting optimization...                                                                  
Optimization finished in 642.3407 seconds with message: Average change in trajectory      
fitness function less than tolerance.                                                     
Results of the analysis are as follows:                                                   
The optimal departure from EARTH occurs at 8/30/2015 14:27:25 (C3=22.1039 km^2/s^2)       
The optimal flyby of VENUS occurs at 2/10/2016 2:20:46 (deltaV=1.9668e-009 km/s, pass     
radius=8185.7625 km)                                                                      
The optimal flyby of VENUS occurs at 3/6/2017 4:56:40 (deltaV=0.0005487 km/s, pass        
radius=12351.9039 km)                                                                     
The optimal flyby of EARTH occurs at 12/31/2018 20:34:37 (deltaV=2.4261e-006 km/s, pass   
radius=8868.4465 km)                                                                      
The optimal arrival at JUPITER BARYCENTER occurs at 1/30/2022 5:50:31 (arrival            
velocity=6.8529 km/s)                                                                     
The optimal trip duration is 2344.641 days.                                               
Generating report file...Done!                                                            
#####################################################################

I did not resort to 'Constrain all flybys to 0 Delta-V'. A low delta V solution should be selected for naturally. Forcing it makes the solver have problems.

I think your problem is in the 'Optimization Options'
Here are my tips:

1) Error Tolerance 1E-2
IMHO, too precise of a tolerance only increases the runtime. You won't fly so precisely in TransX anyway.

2) Population size 750. You need to cast a finer net, so to speak.

3) Max number of iterations 75. If the algorithm doesn't detect the best solution 'family' early on, it probably isn't going to ever. You should know in 75 if you have something worthwhile. This wll still keep the runtime reasonable with the larger population size.

4)Selection Function @selectionstochunif

5) Cost function weights: 2 7 1
You need to put a higher priority on the delta-V cost weight. You can afford a little higher C3 if all of your slingshots are free. Also, I find that low arrival velocity is not mutually exclusive with an overall low delta-V flight plan. If you give that a low priority, it tends to work itself out anyway.

There is an element of chance to genetic algorithms. You won't get the exact same solution twice in a row. Sometimes, it converges on a local min that it happens to detect early-on and it will fail to detect the global min. Run the optimization at least 3 times and see what your best solution is. You can also start to notice families of solutions emerge.
 

downloaderfan

Member
Joined
Jan 12, 2013
Messages
104
Reaction score
0
Points
16
Okay. Actually, you are right and I found a reasonable solution to your EVVEJ problem:

I did not resort to 'Constrain all flybys to 0 Delta-V'. A low delta V solution should be selected for naturally. Forcing it makes the solver have problems.

I think your problem is in the 'Optimization Options'
Here are my tips:

1) Error Tolerance 1E-2
IMHO, too precise of a tolerance only increases the runtime. You won't fly so precisely in TransX anyway.

2) Population size 750. You need to cast a finer net, so to speak.

3) Max number of iterations 75. If the algorithm doesn't detect the best solution 'family' early on, it probably isn't going to ever. You should know in 75 if you have something worthwhile. This wll still keep the runtime reasonable with the larger population size.

4)Selection Function @selectionstochunif

5) Cost function weights: 2 7 1
You need to put a higher priority on the delta-V cost weight. You can afford a little higher C3 if all of your slingshots are free. Also, I find that low arrival velocity is not mutually exclusive with an overall low delta-V flight plan. If you give that a low priority, it tends to work itself out anyway.

There is an element of chance to genetic algorithms. You won't get the exact same solution twice in a row. Sometimes, it converges on a local min that it happens to detect early-on and it will fail to detect the global min. Run the optimization at least 3 times and see what your best solution is. You can also start to notice families of solutions emerge.

Dear boogabooga, again, thanks for your interest in answering my query.

Yes, the solution which you have obtained is a pretty reasonable solution. I was able to obtain the exact same solution using the settings you've mentioned.

But then I added 2 constrains (Total TOF being less than 5 years and the earth venus transfer should take less than 185 days) and again it gave me results with high deltaV.

Here is an image,
a48beaa0da.png


Text data:
Starting optimization...
Optimization finished in 892.0819 seconds with message: Average change in trajectory
fitness function less than tolerance.
Results of the analysis are as follows:
The optimal departure from EARTH occurs at 12/8/2016 11:37:59 (C3=11.788 km^2/s^2)
The optimal flyby of VENUS occurs at 5/8/2017 4:58:40 (deltaV=9.2626 km/s, pass
radius=192500.1537 km)
The optimal flyby of VENUS occurs at 11/3/2017 23:24:51 (deltaV=40.1392 km/s, pass
radius=37844821314918.8 km)
The optimal flyby of EARTH occurs at 12/26/2018 9:10:7 (deltaV=163.3189 km/s, pass
radius=7012.8789 km)
The optimal arrival at JUPITER BARYCENTER occurs at 5/6/2020 13:15:4 (arrival
velocity=14.4967 km/s)
The optimal trip duration is 1245.0674 days.

It seems to me that this tool can definitely produce solutions, but it'll require a lot of tinkering with the 'Optimization options', multiple iterations with the same & different options and a lot of patience of course while it does the calculation.

Anyway, my main point of starting this thread is to better understand this tool and I think now I do understand it's advantages and limitations better.

BTW, I've also uploaded multiple EVVEJ scenarios here, all of which require fuel only to get to venus and have a lower TOF than 2344 days.
 
Last edited:

boogabooga

Bug Crusher
Joined
Apr 16, 2011
Messages
2,999
Reaction score
0
Points
0
Well, the more constraints you add, the less your solution will "fit" compared to the ideal solution. Makes sense.

It seems to me that this tool can definitely produce solutions, but it'll require a lot of tinkering with the 'Optimization options', multiple iterations with the same & different options and a lot of patience of course while it does the calculation.

That's how it is with the Genetic Algorithm, which is a sort of random brute-force technique. I believe that a Design-of-Experiment/Response Surface Methodology approach would be better suited to this problem. It would probably give more consistent results.
 

fausto

FOI SuperMod
Joined
Jul 13, 2008
Messages
797
Reaction score
2
Points
0
Location
Monza (Milan)
I found some years ago a site with a full list of slingshots trajectories for many years to come.. I used them to plan an EVEE cruise to jupiter.. I don t teme ber the url... But you can use Google to find them!
 
Top