[ 3 / biz / cgl / ck / diy / fa / ic / jp / lit / sci / vr / vt ] [ index / top / reports ] [ become a patron ] [ status ]
2023-11: Warosu is now out of extended maintenance.

/sci/ - Science & Math

View post   
File: 52 KB, 652x580, spacex.jpg [View same] [iqdb] [saucenao] [google]
10177515 No.10177515 [Reply] [Original]

How does a rocket do its hoverslam calculations?

>> No.10177552

>>10177515
feedback control

>> No.10177569

calc 1 here, i'm no expert, but i have only a week left in my calc class. I'd probably say it has something to do with the fact, s=1/2(a)t^2+(a)t+So which can be found taking the second integral of acceleration. find out the acceleration and then i would know.

>> No.10177573

>>10177515
It knows where it is, because it knows where it isn't.

>> No.10177619

>>10177573
classic

>> No.10177683

>>10177515
Engie here, just guessing. Most probably by an on-board real-time computational device with feedback loops from sensors, attitude, altitude, GPS, gyro etc...
Same way your drone lands in front of you when you push RTH/RTL two miles out. It knows where to go in what attitude and knows the controls and engines characteristics to efficiently close the gap in a recursive manner.

>> No.10177813

You can do a simple approximation with the rocket equation derived from F = ma, granted to solve it analytically you ignore drag and to set it down lightly you have to be able to engineer a constant acceleration from your thrust.

>> No.10177818

Kalman filters breh

>> No.10177858

https://www.youtube.com/watch?v=c-FT0T9Ei-4

>> No.10178368

Very carefully

>> No.10178382

>>10177683
Yeah, only there's more to it than that because it also needs to know things like how much it weighs, because that then also leads into how much energy it needs to negate to get to stop.
So, yeah, >>10177813 too.

>> No.10178384

>>10177515
Quickly.

>> No.10178402

>>10177515
AFAIK, they use some machine learning algos

>> No.10178701

>>10177515
simple LQG

>> No.10179092
File: 1.56 MB, 1280x720, KerbalSpaeProgram.webm [View same] [iqdb] [saucenao] [google]
10179092

>>10177515
I have nothing useful to add

>> No.10179234

>>10177515
>>>/wsg/2515513
duh

>> No.10179245
File: 15 KB, 251x242, 6dBt2Oj.jpg [View same] [iqdb] [saucenao] [google]
10179245

>>10177515
it practices on your mom

>> No.10179259

SpaceX has a programming team of less than 100 people, really small for the stuff they do. They seem to have taken the Fred Brooks approach to projects.

>> No.10179283

>>10177573
tpbp

>> No.10179284

>>10177858
>>10177573
these /sci/entists know

>> No.10179294

>>10178382
Doesn't really need to know weight. If at current acceleration it calculates it will go positive vertical velocity before touching 0 altitude, it increases thrust. If at max thrust it calculates it won't make 0 vertical before touching 0 altitude, it diverts heading so it crashes into water. No weight necessary, just acceleration, distance till landing, and time until empty. If it knows all engines lit are operating at desired capacity, then it knows thrust (force) and it measures acceleration with accelerometer, so it can derive its own mass

>> No.10179296

>>10179294
Decreases thrust*
The point of hoverslam is kinda you go max thrust all the way until just above surface. If it calculates it will hover too soon, it lowers thrust

>> No.10179312

>>10179259
>100 programmers
> really small for the stuff they do.
nah

>> No.10179579

>>10179296
Scott Manley did a few videos on this stuff. I'm not going to explain to you, go watch some videos.

>> No.10179666

>>10179579
I watch plenty of Scott Manley. If you didn't want to have a conversation why bother replying?

>> No.10179954

>>10177573
came here to post this

>> No.10181037

>>10177515
On-board micro abacus

>> No.10181605

>>10179579
HULLO