Request Interplanetary Superhighway Navigator

[MikeB]

The winner of the Intel Science Talent Search is Erika DeBenedictis. She won for a program that calculates low-energy trajectories through the solar system, much faster than previous approaches have found.

I can't evaluate her claims or her code, but I bet some of our MFD wizards could adapt her approach (assuming it holds up to scrutiny) to a navigational tool for Orbiter.

Her report is at http://www.debenedictis.org/erika/ and is titled "An Optimization Algorithm for Space Mission Design: Dynamically Simulating Energy-Efficient Trajectories"

A video interview is at

Volunteers?

 

[Urwumpe]

I'll at least take a look at it. She has written a lot of stuff about Monte Carlo methods, maybe this one is similar. I wouldn't be surprised if a smart kind of iterative monte carlo solver works faster as concurrent numerical solvers. The question is if the algorithm is still determinative enough and also converges in any situation.

EDIT - yes, she does use Monte Carlo, and she does actually not even use a very smart one, she just sends many virtual spacecraft around and iteratively checks them for death conditions. Using some concepts from modern ray-tracing or genetic algorithms could even improve the performance a lot, this one just searches for all "random" trajectories to arrive at a target. also she uses only Kepler solutions for the positions of the planets, while the Runge-Kutta solver for the spacecraft is equivalent to orbiter.

 

[Sky Captain]

Maybe I`m missing something but I have always thought that traditional Hofman transfer is lowest energy trajectory possible when going for example from Earth to Mars. Or is there a way how to do that with even less energy. But how? you have to perform ejection burn to get free from Earth gravity, and enter solar orbit so your apoapsis touches Mars orbit and then perform a braking burn or aerocapture to enter orbit around Mars. At least it seems only thing you can get for free is aerocapture, but she was talking about using gravity in some way to save fuel. Maybe it was meant to be some elaborate slingshotting maneuvers?

 

[Urwumpe]

No, you can also use the dynamics of the gravity fields in the solar system for making transfers between lagrange points of various planets and moons. This is just much slower than Hohmann.

Imagine it as surfing on waves. Slow waves.

 

[Arrowstar]

Did I read right? She was in high school when she wrote that?

 

[Sky Captain]

But you still have to reach the lagrange points from LEO. IIRC Earth - Sun L2 is ~1.5 million km from Earth so you have to attain nearly 11 km/s to get to it. Now once your spacecraft has arrived at L2 I suppose you wait till the Mars is at suitable position and perform some sort of transfer from Earth L2 to Mars L1 point. Is this how it`s supposed to work?

 

[Urwumpe]

Did I read right? She was in high school when she wrote that?

Yes, and had some lot of help from her Dad. At least the appendix is full of Matlab statements, I can't tell that Matlab was common in my school.

 

[Donamy]

Investing in asteroid mining, may be a good investment, for your grandchildren.