This commit contains a wrong solution for the 0.1 problem before we knew of coth function

- fix the 0.1 speed problem that we went to low, but still havent solve the -0.1 acceleration problem
This commit is contained in:
Marcel Märtens 2022-06-16 21:40:08 +02:00
parent 597fc1b147
commit 3cd99b4711

View File

@ -299,7 +299,7 @@ fn cant_run_during_fall() -> Result<(), Box<dyn Error>> {
#[test] #[test]
fn physics_theory() -> Result<(), Box<dyn Error>> { fn physics_theory() -> Result<(), Box<dyn Error>> {
let tick = |i: usize, move_dir: f64, vel: f64, pos: f64, dt: f64| { let tick = |i: usize, move_dir: f64, acc: f64, vel: f64, pos: f64, dt: f64| {
/* /*
ROLLING_FRICTION_FORCE + AIR_FRICTION_FORCE + TILT_FRICT_FORCE + ACCEL_FORCE = TOTAL_FORCE ROLLING_FRICTION_FORCE + AIR_FRICTION_FORCE + TILT_FRICT_FORCE + ACCEL_FORCE = TOTAL_FORCE
@ -334,17 +334,21 @@ fn physics_theory() -> Result<(), Box<dyn Error>> {
// controller // controller
// I know what you think, wtf, yep: https://math.stackexchange.com/questions/1929436/line-integral-of-force-of-air-resistanc // I know what you think, wtf, yep: https://math.stackexchange.com/questions/1929436/line-integral-of-force-of-air-resistanc
// basically an integral of the air resistance formula which scales with v^2 transformed with an ODE. // basically an integral of the air resistance formula which scales with v^2 transformed with an ODE.
let vel = if acc == 0.0 {
vel // The function besically takes its last result calculates the inverse* of it, adds the newly get speed to it and then run tanh again
} else { // *inverse of the tanh and the factors before.
(mass * acc.abs() / c).sqrt() * ( ( (c / acc.abs() / mass ).sqrt() * old_vel ).atanh() + acc.signum() * (c * acc.abs() / mass).sqrt() * dt ).tanh() let past_fak = (c / (mass * acc.abs() ) ).sqrt() * old_vel;
}; // the original algorithm isn't able to keep a speed over the terminal velocity based on acc. however that is necessary, e.g. for pushbacks, falling, external factors and stopping (because of we stop acc would be 0 and the terminal vel would be 0 too)
// here we decide to reduce that factor by c each second.
let over_vel_keep = (past_fak * c.powf(dt)).max(1.0) * past_fak.signum(); //TODO signum needed
let vel = ( ((mass * acc.abs() ) / c ).sqrt() * over_vel_keep) * ( ( past_fak.clamp(-1.0, 1.0) ).atanh() + acc.signum() * (c * acc.abs() / mass).sqrt() * dt ).tanh();
//let vel = ((mass * acc.abs() + old_acc ) / c).sqrt() * ( ( ( (c / (mass * acc.abs() ) ).sqrt() * old_vel).clamp(-1.0, 1.0) ).atanh() + acc.signum() * (c * acc.abs() / mass).sqrt() * dt ).tanh();
//physics //physics
let distance_last = mass / c * ( ( ( old_vel * (c/acc / mass).sqrt()).atanh() ).cosh() ).ln(); let distance_last = mass / c * ( ( ( old_vel * (c/acc / mass).sqrt()).atanh() ).cosh() ).ln();
let distance = mass / c * ( ( ( old_vel * (c/acc / mass).sqrt()).atanh() + dt* (c * acc / mass).sqrt()).cosh() ).ln(); let distance = mass / c * ( ( ( old_vel * (c/acc / mass).sqrt()).atanh() + dt* (c * acc / mass).sqrt()).cosh() ).ln();
let diff = distance - distance_last; let diff = distance - distance_last;
let diff = if diff.is_nan() { let diff = if !diff.is_finite() {
0.0 0.0
} else { } else {
diff diff
@ -353,10 +357,16 @@ fn physics_theory() -> Result<(), Box<dyn Error>> {
let pos = pos + diff; let pos = pos + diff;
let ending = ((i+1) as f64 *dt * 10.0).round() as i64;
let line = format!("[{:0>2.1}]: move_dir: {:0>1.1}, over_vel_keep: {:0>4.4}, acc: {:0>4.4}, vel: {:0>4.4}, dist: {:0>7.4}, dist: {:0>7.4}, pos: {:0>7.4}, c: {:0>4.4}", (i+1) as f64 *dt, move_dir, over_vel_keep, acc, vel, distance_last, distance, pos, c);
if ending % 10 == 0 {
println!("\x1b[91m{}\x1b[0m", line)
} else if ending % 2 != 0 {
println!("\x1b[94m{}\x1b[0m", line)
} else {
println!("{}", line)
}
//if ((i+1) as f64 *dt * 10.0).round() as i64 % 2 == 0 {
println!("[{:0>2.1}]: move_dir: {:0>1.1}, acc: {:0>4.4}, vel: {:0>4.4}, dist: {:0>7.4}, dist: {:0>7.4}, pos: {:0>7.4}, c: {:0>4.4}", (i+1) as f64 *dt, move_dir, acc, vel, distance_last, distance, pos, c);
//}
(acc, vel, pos) (acc, vel, pos)
}; };
@ -364,18 +374,18 @@ fn physics_theory() -> Result<(), Box<dyn Error>> {
let dt = 1.0 / tps as f64; let dt = 1.0 / tps as f64;
println!(""); println!("");
println!("dt: {}", dt); println!("dt: {}", dt);
let (_acc, mut vel, mut pos) = (0.0, 0.0, 0.0); let (mut acc, mut vel, mut pos) = (0.0, 0.0, 0.0);
let mut i = 0; let mut i = 0;
for _ in 0..tps { for _ in 0..tps*2 {
(_, vel, pos) = tick(i, 1.0, vel, pos, dt); (acc, vel, pos) = tick(i, 1.0, acc, vel, pos, dt);
i += 1;
}
for _ in 0..tps*2 {
(acc, vel, pos) = tick(i, 0.1, acc, vel, pos, dt);
i += 1; i += 1;
} }
for _ in 0..tps { for _ in 0..tps {
(_, vel, pos) = tick(i, -1.0, vel, pos, dt); (acc, vel, pos) = tick(i, 1.0, acc, vel, pos, dt);
i += 1;
}
for _ in 0..tps {
(_, vel, pos) = tick(i, 1.0, vel, pos, dt);
i += 1; i += 1;
} }
(vel, pos) (vel, pos)