How precise are our orbit/intercept calculations vs the ksp?

[Tripzter]

So something i've always wondered ever since i started playing ksp. How exact is our current technology for determining the orbit/planet encounter altitude? For example orbit altitude levels as well as docking with other targets in space... how about when going to another planet?

[Sirine]

This is a game.

Try out the RSS (Real Solar System). With couple of 'try to be real' mods. will get you basic ideal of how much the difference.

Long words short. You can't fly a real "rocket", "plane", via playing this games. This is not a simulation base program. It just a game.

Edited August 4, 2014 by Sirine

[LethalDose]

As Sirine pointed out, the game and real life are barely comparable.

That being said, IRL astro navigation is really really good, and it has to be is really hard. Back when Scott Manley was doing "Deep Space Hangouts" on G+, there was a guy from JPL he had on a few times. From the sound of it, patched comics work great in real life, but due non-gravitational effects, simulation and numerical methods are still the best methods for estimating trajectories. Additionally, methods like laser range finding, radar altimeter, and stellar navigation (using star positions) vastly improve the ability of probes to know where they are in space

EDIT: it's also worth noting that the Russians have been using automated docking systems (Igla and Kurs) for docking the Soyuz and Progress spacecraft for almost 50 years, so that's gotten pretty effective too.

Edited August 4, 2014 by LethalDose

[Grover]

Even the Apollo spacecraft were capable of tracking their position relative to earth in real-time with assistance from ground stations (untill you accidentally reboot the navigation system in flight...)

Nowadays, its a combination of loads of complex systems, most notably a reverse GPS system, which uses foxed ground stations to track a moving satalite, rather than a fixed (orbit) satalite tracking a ground vehicle.

[Tripzter]

This is a game.

Try out the RSS (Real Solar System). With couple of 'try to be real' mods. will get you basic ideal of how much the difference.

Long words short. You can't fly a real "rocket", "plane", via playing this games. This is not a simulation base program. It just a game.

You might actually want to read what i wrote. I'm simply asking how exact our real life position/trajectory trackers are vs the ksp... I'm not trying to fly a ship in ksp and pretend that i can fly nasas shuttle to the mars and back in a blink of an eye...

[Maxwell Fern]

IRL caclculations higly depend on the distance and the size of the object. While it is possible to know where a satellite in low Earth orbit is to the centimeter, it's a whole different story to know where is a distant asteroid and where it's going. It takes several observations over a long period of time to be able to say "it has one chance in a million to hit earth in the first semeceter of 2075". Which is a fancy way to say "yeah, it's getting closer to our planet but its trajectory shouldn't cross ours, although we might be wrong, we'll check again next year".

Edited August 7, 2014 by Maxwell Fern

[Linear]

You might actually want to read what i wrote. I'm simply asking how exact our real life position/trajectory trackers are vs the ksp... I'm not trying to fly a ship in ksp and pretend that i can fly nasas shuttle to the mars and back in a blink of an eye...

He probably also means that the current SOI based physics used in ksp are just a rough model, it doesn't actually work like that.

[problemecium]

Indeed - the "encounter altitude" or any other sort of discrete moment of encounter is at best a simplification in real life. Even on the ground on Earth, for example, the Moon's gravity has an effect on the trajectories of objects, so as a craft flies in space it actually gradually shifts from being mostly influenced by one thing to being more heavily influenced by something else. Hence patched conics not being completely accurate for many real-life missions.

[ThinkOutsideTheHangar]

Its close enough. And besides, n-body physics would make me barely able to maintain an orbit.

[Lysergic]

NASA is really really good at their intercept/orbit calculations. They can fling a probe to mars and have it get into the orbit they predicted give or take a mile or two. It's remarkable.

Basically NASA has the tools to be a lot more precise and accurate than you can be in KSP.

[Renegrade]

NASA is really really good at their intercept/orbit calculations. They can fling a probe to mars and have it get into the orbit they predicted give or take a mile or two. It's remarkable.

A mile or two, eh?

[Lysergic]

Yeah, well that's the equivalent of forgetting solar panels or something. Everyone has brain farts.

[R0cketC0der]

Actually patched conics was used by NASA during the Apollo program so I consider it close enough.

[Armchair Rocket Scientist]

I seem to remember that one interplanetary spacecraft, not sure which one, arrived at its destination within a few seconds of the predicted time.

[Nibb31]

Real-life orbit predictions are extremely accurate because they very rarely miss their target. But their calculations can be immensely complex, especially when dealing with interplanetary transfers, slingshots, halo orbits, and such.

Now, imagine the guy who had to come up with that trajectory in the 70's when asked how to send a probe to visit Halley's Comet.

KSP orbit predictions are 100% accurate, because the simplified predictions use the same simplified formulae as the simplified simulation model. If I create a mathematical model with the formula "1+1=2", then I can use the same formula to predict the result with 100% accuracy.

So to answer the OP's post, KSP orbital predictions are more precise than real-life space agency predictions.

Edited August 8, 2014 by Nibb31

[Renegrade]

Yeah, well that's the equivalent of forgetting solar panels or something. Everyone has brain farts.

Yeah, I know. I just couldn't help poking fun at the Mars/Mile thing. Sorry

I don't think I've ever forgotten to install solar panels, by the way. It's the opening of said panels that I forget.

[YNM]

The RL approximation are far more accurate even with the higher complexity. I've sent an ion-powered spacecraft to Kerbol, and it's Pe is decreasing a meter each second because of rounding errors. Whilst IRL, the computers used are far more superior than our personal computers for simulations and computations. Just think it this way - even the Sun's light are actually pushing interplanetary probe further - you could think what's the accuracy required to take that into account, so the probe can aerobrake instead of lithobrake. Or say, for L points, you must take the perturbation of other bodies into account so you can pack the right amount of propellants for stationkeeping throughout the mission time.

The only times there were errors are due to simplications, or due to unit mismatch.