Overview Evidence tour Methodology Dead zones Problems

2D APN with Evasive Target

ADVANTAGES1 · dim 6

SolvSRK wins. At the comparison noise level, SolvSRK beats the best baseline by at least 10 percentage points of survival, or by at least 0.05 balanced score when survival is tied. Use SolvSRK for this class of problem. All verdicts →

2D augmented PN (N_eff=4) against sinusoidal-weave target (50 m/s^2 at 0.5 Hz). Maneuver switching creates moderate stiffness.

Guidance & interception

Problem definition

Zarchan Ch. 8 (augmented PN with target maneuver)

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 _clamp(x: float, lo: float, hi: float) -> float:
    return max(lo, min(hi, x))

def rhs(t, y):
    rx, ry = y[0], y[1]
    vx_i, vy_i = y[2], y[3]
    vx_t, vy_t = y[4], y[5]

    d = np.empty(6)

    R = np.sqrt(rx * rx + ry * ry)
    if R < _RANGE_GUARD:
        d[:] = 0.0
        return d

    lam = np.arctan2(ry, rx)

    vx_rel = vx_i - vx_t
    vy_rel = vy_i - vy_t

    R_dot = (rx * vx_rel + ry * vy_rel) / R
    lam_dot = (rx * vy_rel - ry * vx_rel) / (R * R)
    lam_dot = _clamp(lam_dot, -_LOS_RATE_CLAMP, _LOS_RATE_CLAMP)

    V_c = -R_dot

    a_tgt_now = A_tgt * np.sin(omega_evade * t)
    a_cmd = N_eff * V_c * lam_dot + 0.5 * N_eff * a_tgt_now

    perp_x = -np.sin(lam)
    perp_y = np.cos(lam)

    wx = np.interp(t, _noise_ts, wind_noise[0])
    wy = np.interp(t, _noise_ts, wind_noise[1])

    d[0] = vx_i - vx_t
    d[1] = vy_i - vy_t
    d[2] = a_cmd * perp_x + wx
    d[3] = a_cmd * perp_y + wy
    d[4] = -a_tgt_now * np.sin(lam)
    d[5] = a_tgt_now * np.cos(lam)

    return d

Parameters: A_tgt = 50
N_eff = 4
_LOS_RATE_CLAMP = 0.5
_RANGE_GUARD = 0.1
_noise_ts = [0, 0.01, 0.02, 0.03, 0.04, 0.05, …] [shape=(2501,), min=0, max=25]
omega_evade = 3.14159265359
wind_noise = [0.0628651105467, -0.0660524316457, 0.320211325222, 0.0524500585765, -0.267834686581, 0.180797527455, …] [shape=(2, 2501), min=-1.94971086503, max=1.62859953736]
Initial condition: y(0) = [5000, 2000, -250, 0, -200, 0]
Horizon: t ∈ [0, 25]

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: none

Recommendation snapshot

Clean best: SolvSRK

Noisy best: SolvSRK

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: 2D APN with Evasive Target (2d-apn-with-evasive-target)

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

#	Solver	Survival	SCD	nfev	Wall	Score
1	SolvSRK	100%	9.2	204,474	1.86 s	0.838
2	SciPy LSODASciPy	100%	6.6	80,862	744 ms	0.775
3	FBDFexternal	100%	6.3	101,967	7.34 s	0.769
4	CVODE BDFexternal	100%	6.0	57,360	708 ms	0.761
5	SciPy RK23SciPy	100%	5.9	62,189	857 ms	0.759
6	SciPy BDFSciPy	100%	5.9	67,986	2.81 s	0.759
7	CVODE Adamsexternal	100%	5.9	45,916	563 ms	0.759
8	Vern9external	100%	5.8	224,018	8.82 s	0.757
9	Tsit5external	100%	5.6	84,120	8.23 s	0.753
10	SciPy RadauSciPy	100%	5.3	172,097	4.70 s	0.745
11	Vern7external	100%	5.3	165,562	7.44 s	0.744
12	SciPy DOP853SciPy	100%	5.2	209,954	2.41 s	0.742
13	SciPy RK45SciPy	100%	5.1	56,720	681 ms	0.742
14	TRBDF2external	100%	4.1	30,774	5.82 s	0.717

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 →

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_2d_apn_with_evasive_target_2026,
  title        = {Resonix Evidence Portal: 2D APN with Evasive Target},
  author       = {{Resonix Labs (Canada) Inc.}},
  year         = {2026},
  howpublished = {\url{https://resonixusa.com/evidence/problems/2d-apn-with-evasive-target}},
  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.}
}

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

Problem definition

Versions & freeze

Results matrix

Cite this page

Related