Skip to content

PBFE Numerical Illustration v1.1.0 — UQ and Target-Oriented Extension

Choose a tag to compare

View all tags
@hufanyoung hufanyoung released this 11 Aug 08:12
· 2 commits to main since this release
v1.1.0
03127e5

This release extends the reproducible reduced-form numerical illustration for the Performance-Based Food Engineering (PBFE) framework.

New in v1.1.0

  • Continuous Monte Carlo verification of the synthetic conditional hierarchy.
  • First-order reliability method (FORM) approximation at the illustrative DV = 1 limit.
  • Local design-point directional-cosine diagnostics.
  • Six one-dimensional synthetic parameter-to-target scenarios.
  • A manuscript-facing target-oriented assessment figure.
  • Machine-readable diagnostics, sensitivity results, and target profiles.
  • A presentation-only update to the loss-exceedance figure that adds a small lower-axis margin without changing the numerical results.

The authoritative direct discrete exceedance probability remains:

P(DV > 1) = 0.09591450211754649

Continuous Monte Carlo verification gives:

  • P(DV > 1) = 0.095851
  • standard error = 0.0002943868
  • N = 1,000,000
  • seed = 20260810

The FORM approximation gives:

  • beta = 1.22556
  • P_FORM = 0.110182

For the illustrative 5% probability target, resolved one-dimensional parameter crossings are obtained for the LAI nominal mean, TFP-link nominal mean, normalized-loss nominal mean, and normalized-loss COV scenarios. No 5% crossing was resolved for the LAI-COV or TFP-COV scenario in the extended numerical search.

All numerical relationships and parameter scenarios are synthetic. They do not represent calibrated agricultural interventions, empirical economic models, or validated strategy rankings.

Validation

  • Baseline tests: 5/5 passed
  • UQ/target extension tests: 16/16 passed
  • Combined tests: 21/21 passed
Assets 2
Loading