Certified Optimal Adaptive Recovery in a Finite Partially Observed Controlled System: A 24-Phase Fibonacci-Modulo-9 Case Study

We present a fully specified finite partially observed adversarial control model with 24 phases, action-dependent transitions, Fibonacci-modulo-9 observation channels, and a safe reachability objective. For the initial belief B₀ = {1, 22}, disturbance set Δ = {0, 4}, and finite horizon H = 5, the optimal worst-case cost under adaptive observation selection is exactly 10. An explicit six-record strategy certificate establishes the upper bound, while an independently implemented exhaustive Bellman algorithm establishes the matching lower bound. Under the same cost convention, fixed local, external, and joint observation regimes have optimal worst-case costs of 12, 15, and 19, respectively. Thus, adaptive observation selection achieves a strict cost advantage in the specified finite instance. The study additionally includes an exact robust vector-Pareto analysis for horizon three, a separate greatest-fixed-point analysis of infinite-horizon safety, independent verification routines, and reproducible Python source code using only the standard library. The contribution is a certified, reproducible finite mathematical case study. No claim is made of a new general belief-state algorithm, an experimentally established physical Q-field, or independently verified historical novelty. The manuscript is a preprint and has not undergone independent peer review. This preprint presents a finite, computer-assisted mathematical case study within the author's Reconstruction Theory research program. The accompanying Python implementation is designed for offline execution using the Python standard library, including on Android through Pydroid 3. The computational package contains a fixed strategy certificate, an independent minimax lower-bound verification, robust Pareto calculations, and an infinite-horizon safety analysis. AI assistance was used in drafting, code development, mathematical exposition, and preliminary consistency checks. The author reviewed and approved the release and retains responsibility for its contents. The manuscript has not undergone independent peer review. The results concern a stipulated finite mathematical model and do not establish a physical Q-field.

Authors

Publication Details

Journal
Zenodo (CERN European Organization for Nuclear Research)
Published
2026-11-18
DOI
https://doi.org/10.5281/zenodo.24351095
Primary Topic
Formal Methods in Verification
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

Certified Optimal Adaptive Recovery in a Finite Partially Observed Controlled System: A 24-Phase Fibonacci-Modulo-9 Case Study

Ivo Vajc
Zenodo (CERN European Organization for Nuclear Research)
Formal Methods in Verification
preprint

Certified Optimal Adaptive Recovery in a Finite Partially Observed Controlled System: A 24-Phase Fibonacci-Modulo-9 Case Study

Ivo Vajc
preprint en

Abstract

We present a fully specified finite partially observed adversarial control model with 24 phases, action-dependent transitions, Fibonacci-modulo-9 observation channels, and a safe reachability objective. For the initial belief B₀ = {1, 22}, disturbance set Δ = {0, 4}, and finite horizon H = 5, the optimal worst-case cost under adaptive observation selection is exactly 10. An explicit six-record strategy certificate establishes the upper bound, while an independently implemented exhaustive Bellman algorithm establishes the matching lower bound. Under the same cost convention, fixed local, external, and joint observation regimes have optimal worst-case costs of 12, 15, and 19, respectively. Thus, adaptive observation selection achieves a strict cost advantage in the specified finite instance. The study additionally includes an exact robust vector-Pareto analysis for horizon three, a separate greatest-fixed-point analysis of infinite-horizon safety, independent verification routines, and reproducible Python source code using only the standard library. The contribution is a certified, reproducible finite mathematical case study. No claim is made of a new general belief-state algorithm, an experimentally established physical Q-field, or independently verified historical novelty. The manuscript is a preprint and has not undergone independent peer review. This preprint presents a finite, computer-assisted mathematical case study within the author's Reconstruction Theory research program. The accompanying Python implementation is designed for offline execution using the Python standard library, including on Android through Pydroid 3. The computational package contains a fixed strategy certificate, an independent minimax lower-bound verification, robust Pareto calculations, and an infinite-horizon safety analysis. AI assistance was used in drafting, code development, mathematical exposition, and preliminary consistency checks. The author reviewed and approved the release and retains responsibility for its contents. The manuscript has not undergone independent peer review. The results concern a stipulated finite mathematical model and do not establish a physical Q-field.

Zenodo (CERN European Organization for Nuclear Research)
Formal Methods in Verification
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon