Files
2026-02-16 21:52:26 +08:00

35 lines
1.1 KiB
Markdown
Raw Permalink Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
TASK: UQ for TTE by stochastic usage paths; output CI + survival curve.
INPUT DATA:
- z0 = 1.00
- Baseline params
- Base deterministic inputs L0(t), C0(t), N0(t) from scenario
- Stochastic perturbations: OU processes added to each of L,C,N:
dX = -θ(X-0)dt + σ dW
Use θ=1/600 1/s (10-minute reversion), σ=0.02
- Enforce bounds by clipping final L,C,N to [0,1]
- Runs:
M = 300 Monte Carlo paths
seed = 20260201
METHODOLOGY:
1) For m=1..M, generate OU noise paths on the same dt grid.
2) Build L_m(t)=clip(L0(t)+X_L(t)), etc.
3) Simulate → TTE_m.
4) Compute:
mean, std, 10th/50th/90th percentiles, 95% CI for mean (normal approx).
5) Survival curve:
For t_grid_hours = 0..max(TTE) in 0.25h increments,
estimate S(t)=P(TTE > t) empirically.
DELIVERABLES:
A) UQ_SUMMARY_v1 (JSON): mean, std, p10, p50, p90, CI95_low, CI95_high
B) SURVIVAL_CURVE_v1 (CSV): t_hours, S(t)
C) REPRODUCIBILITY_v1 (JSON): seed, M, θ, σ, dt
VALIDATION:
- Must have exactly M successful runs.
- Survival curve must be non-increasing in t (else FAIL).
OUTPUT FORMAT:
JSON, CSV, JSON. No prose.