ESC Dynamics with Thermal Derating (8-state)

PARITYS3 · dim 8

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 →

8-state ESC dynamics: output voltage/current, MOSFET/board temperatures, duty cycle, DC bus capacitor, cumulative energy. PWM switching tau ~0.001s vs thermal tau ~60s → stiffness ~60000:1.

Drone dynamics & autonomy

Problem definition

Mohan, Undeland & Robbins, 'Power Electronics'; STMicroelectronics AN4070

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 _duty_command(t: float) -> float:
    """Simulated flight controller PWM commands."""
    if t < 0.5:
        return 0.3          # idle / spin-up
    elif t < 2.0:
        return 0.65         # hover throttle
    elif t < 4.0:
        return 0.85         # climb / aggressive manoeuvre
    elif t < 6.0:
        return 0.65         # back to hover
    elif t < 8.0:
        return 0.4          # descent
    else:
        return 0.1          # motor brake / idle

def _motor_omega(duty: float) -> float:
    """Approximate motor speed from duty cycle for back-EMF."""
    return duty * _V_BUS / _KE * 0.85

def rhs_esc_dynamics(t, y):
    """8-state ESC dynamics with current limiting and thermal derating."""
    dy = np.zeros(8)

    V_out = y[0]
    I_out = y[1]
    T_mos = y[2]
    T_brd = y[3]
    duty = y[4]
    I_cap = y[5]
    V_cap = y[6]

    # PWM filter: track commanded duty with first-order lag
    d_cmd = _duty_command(t)

    # Thermal derating
    if T_mos > _T_DERATE:
        derate = max(0.0, 1.0 - (T_mos - _T_DERATE) / (_T_SHUTDOWN - _T_DERATE))
    else:
        derate = 1.0
    d_limited = d_cmd * derate

    # Current limiting (soft)
    if I_out > _I_SOFT:
        i_limit_factor = max(0.1, 1.0 - (I_out - _I_SOFT) / (_I_MAX - _I_SOFT))
        d_limited *= i_limit_factor

    dy[4] = (d_limited - duty) / _TAU_PWM

    # Output voltage
    omega = _motor_omega(duty)
    V_bemf = _KE * omega
    V_applied = V_cap * duty

    dy[0] = (V_applied - V_out) / (_TAU_PWM * 10)

    # Output current: V = I*R + L*dI/dt + V_bemf
    if _L_MOTOR > 0:
        dy[1] = (V_out - I_out * _R_MOTOR - V_bemf) / _L_MOTOR
    else:
        dy[1] = 0.0

    # Clamp current
    if I_out >= _I_MAX and dy[1] > 0:
        dy[1] = 0.0

    # MOSFET conduction + switching losses
    P_cond = I_out**2 * _R_DS_ON * 2  # 2 FETs in current path
    P_sw = 0.5 * V_cap * I_out * 20e-9 * _F_SW * 2  # switching loss estimate
    P_total = P_cond + P_sw

    # Thermal dynamics: junction
    dy[2] = (P_total - (T_mos - T_brd) / _R_TH_JC) / _C_TH_J

    # Thermal dynamics: board
    dy[3] = ((T_mos - T_brd) / _R_TH_JC - (T_brd - _T_AMB) / _R_TH_BA) / _C_TH_B

    # DC bus capacitor
    I_bus = I_out * duty
    dy[5] = (I_bus - I_cap) / (_TAU_PWM * 5)
    dy[6] = (I_cap - I_bus) / _C_BUS if _C_BUS > 0 else 0.0

    # Cumulative energy dissipated
    dy[7] = P_total

    return dy
Parameters
  • _C_BUS = 0.00047
  • _C_TH_B = 15
  • _C_TH_J = 0.5
  • _F_SW = 16000
  • _I_MAX = 30
  • _I_SOFT = 25
  • _KE = 0.007
  • _L_MOTOR = 0.0001
  • _R_DS_ON = 0.008
  • _R_MOTOR = 0.1
  • _R_TH_BA = 20
  • _R_TH_JC = 3
  • _TAU_PWM = 6.25e-05
  • _T_AMB = 25
  • _T_DERATE = 100
  • _T_SHUTDOWN = 150
  • _V_BUS = 14.8
Initial condition
y(0) = [0, 0, 25, 25, 0, 0, 14.8, 0]
Horizon
t ∈ [0, 10]

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: extreme

Default noise: low

Recommendation snapshot

Clean best: SciPy BDF

Noisy best: SciPy RK45

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: ESC Dynamics with Thermal Derating (8-state) (esc-dynamics-with-thermal-derating-8-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
1SciPy BDFSciPy
100%
8,420266 ms0.809
2SciPy RadauSciPy
100%
22,660467 ms0.809
3SciPy RK45SciPy
100%
95,450793 ms0.809
4SciPy LSODASciPy
100%
8,55945 ms0.809
5SciPy DOP853SciPy
100%
304,7302.34 s0.809
6SciPy RK23SciPy
100%
73,943741 ms0.809
7CVODE BDFexternal
100%
8,70980 ms0.809
8CVODE Adamsexternal
100%
5,45352 ms0.809
9Tsit5external
100%
98,1428.61 s0.809
10SolvSRK
100%
20,12598 ms0.809

At Clean, best balanced arm is SciPy BDF · SolvSRK survival 100%.

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_esc_dynamics_with_thermal_derating_8_state_2026,
  title        = {Resonix Evidence Portal: ESC Dynamics with Thermal Derating (8-state)},
  author       = {{Resonix Labs (Canada) Inc.}},
  year         = {2026},
  howpublished = {\url{https://resonix.tech/evidence/problems/esc-dynamics-with-thermal-derating-8-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