6DOF Quadrotor Rigid Body (12-state)

PARITYS2 · dim 12

No clear winner. The survival gap is under 10 percentage points and the balanced-score gap is under 0.05, so neither SolvSRK nor the best baseline clears the win threshold. Either works — choose on cost, licensing, or integration effort. All verdicts →

12-state rigid body 6DOF quadrotor: position (x,y,z), velocity (u,v,w), Euler angles (phi,theta,psi), angular rates (p,q,r). X-frame motor layout with aero drag. Stiffness from fast motor response vs slow position dynamics.

Drone dynamics & autonomy

Problem definition

Quan, 'Introduction to Multicopter Design and Control' (2017); Beard & McLain, 'Small Unmanned Aircraft'

Canonical RHS excerpt from the registered callable used for this benchmark cell. Expand it to verify the state equations; it is not a standalone runnable fixture.

Show canonical RHS excerpt
def _motor_speeds(t: float) -> np.ndarray:
    """Motor angular velocities for a typical test manoeuvre."""
    w_hover = math.sqrt(_MASS * _G / (4.0 * _Kf))

    if t < 2.0:
        # Hover
        return np.array([w_hover, w_hover, w_hover, w_hover])
    elif t < 5.0:
        # Pitch forward: increase rear, decrease front
        delta = 0.03 * w_hover
        return np.array([w_hover - delta, w_hover + delta,
                         w_hover - delta, w_hover + delta])
    elif t < 8.0:
        # Yaw right: differential CW/CCW
        delta = 0.02 * w_hover
        return np.array([w_hover + delta, w_hover + delta,
                         w_hover - delta, w_hover - delta])
    elif t < 12.0:
        # Return to hover
        return np.array([w_hover, w_hover, w_hover, w_hover])
    else:
        # Slow descent
        return np.array([w_hover * 0.92, w_hover * 0.92,
                         w_hover * 0.92, w_hover * 0.92])

def rhs_6dof_quad(t, y):
    """12-state rigid-body 6DOF quadrotor dynamics.

    State vector: [x, y, z, u, v, w, phi, theta, psi, p, q, r]
    """
    dy = np.zeros(12)

    u, v, w = y[3], y[4], y[5]
    phi, theta, psi = y[6], y[7], y[8]
    p, q, r = y[9], y[10], y[11]

    # Motor forces and torques
    omega = _motor_speeds(t)
    F = _Kf * omega**2
    M = _Km * omega**2

    T_total = np.sum(F)

    # Torques from X-frame motor layout
    L_roll = _L * (-F[0] + F[1] + F[2] - F[3])
    M_pitch = _L * (-F[0] - F[1] + F[2] + F[3])
    N_yaw = np.sum(_SPIN_DIRS * M)

    # Aero drag (body-frame, linear approximation)
    V_body = math.sqrt(u**2 + v**2 + w**2)
    if V_body > 0.01:
        Fd_x = -_CD_BODY * 0.5 * 1.225 * V_body * u
        Fd_y = -_CD_BODY * 0.5 * 1.225 * V_body * v
        Fd_z = -_CD_BODY * 0.5 * 1.225 * V_body * w
    else:
        Fd_x = Fd_y = Fd_z = 0.0

    # Rotation matrix elements (ZYX convention: psi, theta, phi)
    cphi, sphi = math.cos(phi), math.sin(phi)
    cth, sth = math.cos(theta), math.sin(theta)
    cpsi, spsi = math.cos(psi), math.sin(psi)

    # Position derivatives (NED frame)
    dy[0] = (cth * cpsi) * u + (sphi * sth * cpsi - cphi * spsi) * v + \
            (cphi * sth * cpsi + sphi * spsi) * w
    dy[1] = (cth * spsi) * u + (sphi * sth * spsi + cphi * cpsi) * v + \
            (cphi * sth * spsi - sphi * cpsi) * w
    dy[2] = (-sth) * u + (sphi * cth) * v + (cphi * cth) * w

    # Velocity derivatives (body frame, Newton's 2nd law)
    dy[3] = (r * v - q * w) + Fd_x / _MASS - _G * sth
    dy[4] = (p * w - r * u) + Fd_y / _MASS + _G * cth * sphi
    dy[5] = (q * u - p * v) + Fd_z / _MASS + _G * cth * cphi - T_total / _MASS

    # Euler angle derivatives (kinematic equations)
    if abs(cth) > 1e-6:
        dy[6] = p + (sphi * q + cphi * r) * sth / cth
        dy[7] = cphi * q - sphi * r
        dy[8] = (sphi * q + cphi * r) / cth
    else:
        dy[6] = p
        dy[7] = cphi * q - sphi * r
        dy[8] = 0.0

    # Angular rate derivatives (Euler's equations)
    dy[9] = (L_roll - ((_Izz - _Iyy) * q * r)) / _Ixx
    dy[10] = (M_pitch - ((_Ixx - _Izz) * p * r)) / _Iyy
    dy[11] = (N_yaw - ((_Iyy - _Ixx) * p * q)) / _Izz

    return dy
Parameters
  • _CD_BODY = 0.25
  • _G = 9.80665
  • _Ixx = 0.0082
  • _Iyy = 0.0082
  • _Izz = 0.0148
  • _Kf = 0.000611
  • _Km = 1.5e-05
  • _L = 0.159099025767
  • _MASS = 1.5
  • _SPIN_DIRS = [1, 1, -1, -1]
Initial condition
y(0) = [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
Horizon
t ∈ [0, 15]

Canonical RHS excerpt captured from the same registered callable used for the published benchmark. Frozen closure values are summarized below; helper imports and solver settings are intentionally omitted.

Fingerprint

Spread: low

Default noise: medium

Recommendation snapshot

Clean best: SolvSRK

Noisy best: SciPy LSODA

Coverage

14 solver arms · clean + 5 noise levels

Ranked on survival, precision, and speed

Versions & freeze

Methodology →
Freeze
2026-08-13
libsolvsrk
2.3.0
SciPy
1.14
SUNDIALS
CVODE (bundled backend)

20 seeds/cell default · 14 arms · TRL 4–5 · simulation-lab validated · this page: 6DOF Quadrotor Rigid Body (12-state) (6dof-quadrotor-rigid-body-12-state)

Governed SolvTune benchmark freeze; per-arm medians only. RHS definitions and raw trial rows are not published.

Self-reported by Resonix Labs · not independently verified

Results matrix

Pick an objective and a noise level to rank all arms on survival, median SCD, median nfev, and median wall time. Medians across seeds.

Objective

Best overall trade-off of survival, precision, and speed.

Noise level

#SolverSurvivalSCDnfevWallScore
1SolvSRK
100%
9.531,849333 ms0.845
2SciPy RadauSciPy
100%
7.550,2351.67 s0.798
3SciPy LSODASciPy
100%
6.87,862129 ms0.780
4SciPy BDFSciPy
100%
6.317,131774 ms0.769
5SciPy RK23SciPy
100%
5.974,8791.65 s0.760
6CVODE Adamsexternal
100%
5.66,076127 ms0.753
7SciPy DOP853SciPy
100%
5.69,518179 ms0.752
8CVODE BDFexternal
100%
5.27,528158 ms0.742
9SciPy RK45SciPy
100%
5.113,094261 ms0.740
10Tsit5external
100%
4.813,5362.28 s0.733

At Clean, best balanced arm is SolvSRK.

Values are medians across seeds, measured by Resonix Labs on Resonix hardware and not independently verified; nfev and wall are on reference lab hardware (indicative). Under injected noise only SolvSRK and the SciPy arms are run. How we measure accuracy → · Verification status →

SolvScout · free

Profile your problem for free

This page shows one published benchmark cell. SolvScout fingerprints your ODE, compares it to the full corpus, and recommends a solver with the same survival / precision / speed ranking you see here — including when a SciPy arm wins.

SolvSRK · 30-day trial

Run the winner on your machine

SolvSRK is the stiffness-adaptive integrator behind the SolvSRK column in these tables. Create an account, activate a machine, and take a 30-day trial — same binary you'd ship after purchase.

Cite this page

Replace the access date. Pin the freeze ID and library versions when comparing against a later export. Cite it as what it is — a self-reported vendor benchmark, not an independently verified result. The note field says so; please keep it.

@misc{resonix_evidence_6dof_quadrotor_rigid_body_12_state_2026,
  title        = {Resonix Evidence Portal: 6DOF Quadrotor Rigid Body (12-state)},
  author       = {{Resonix Labs (Canada) Inc.}},
  year         = {2026},
  howpublished = {\url{https://resonix.tech/evidence/problems/6dof-quadrotor-rigid-body-12-state}},
  note         = {Self-reported vendor benchmark; internally generated by Resonix Labs and not independently verified. Accessed YYYY-MM-DD. Freeze 2026-08-13; libsolvsrk 2.3.0; SciPy 1.14.}
}

Related

TRL 4–5 · simulation-lab validated · 398 problems · 14 solver arms · clean + 5 noise levels

Freeze: 2026-08-13 · scipy 1.14 · libsolvsrk 2.3.0 · Methodology

Self-reported by Resonix Labs · not independently verified · Verification status