Adverse-Integrity Review: "Distributionally Robust Optimal Scheduling With Heterogeneous Uncertainty Information: A Framework for Hydrogen Systems" (DOI: 10.1109/TSTE.2024.3388388)

Academic fraud report · Geng Detector

Summary

Verdict: Highly suspicious; recommend in-depth investigation. The review of the 2024 IEEE Transactions on Sustainable Energy paper by A. Zhou, M. E. Khodayar and J. Hui Wang identified four concerns. (1) A hard symbolic/logical inconsistency between the multi-index uncertainty set defined in Eq. (8c) (indexed by k = 1,…,K) and the single-scalar reformulation of Ω_t in Eqs. (13)–(17), strongly suggesting copy-paste from a multi-dimensional framework that was never re-aligned with the body’s scalar treatment. (2) Reporting that a 2024 submission was implemented in MATLAB 2014a on 8 GB RAM, atypical for a contemporary MISOCP study on a 118-bus system. (3) Participation factors in Table III cluster suspiciously on a 0.1 grid (0, 0.1, …, 1.0), unlikely to arise unmodified from a continuous MISOCP solver. (4) The proof of Theorem 1 is delegated entirely to the authors’ own companion paper [31], consistent with salami-slicing. Confidence is moderate; image and statistical forensics were not applicable as raw data and figures were not available.

Verdict

🟠 Highly suspicious. Recommended for in-depth institutional review.

Key findings

Symbolic–mathematical contradiction (hard issue): Eq. (8c) defines a moment-based ambiguity set indexed by k = 1,…,K (multivariate uncertainty ξ_k with bounds σ̄²_k ≤ E_P[ξ_k²] ≤ σ̄²_k), whereas the objective reformulation in Section III.A (Eqs. 13–17) treats uncertainty as the scalar Ω_t, using only μ_t and σ_t² with no index k. This mismatch — multidimensional ambiguity set versus scalar expectation — is not a typographical fix; it indicates the multivariate formulation was transplanted from another work without reconciling the scalar derivation in the body.
Anachronistic experimental environment (moderate concern): Section IV states simulations were performed in MATLAB 2014a on an Intel Core i5 with 8 GB RAM. The paper was submitted in October 2023 and accepted in early 2024; contemporary MISOCP solvers (Gurobi, CPLEX) and modern MATLAB releases (R2023a/b) are standard. Reporting a 20-second solve time on such legacy infrastructure for a 118-bus hydrogen-coupled system is implausible.
Overly idealized participation factors (moderate concern): Table III values cluster on a 0.1 grid (0, 0.1, 0.2, 0.8, 0.9, 1.0). Continuous MISOCP solvers typically return non-rounded values (e.g., 0.1345, 0.8655); perfectly gridded outputs across all 24 hours suggest post-hoc rounding or model oversimplification.
Self-citation / salami-slicing pattern (low–moderate concern): Theorem 1 proof is delegated in full to "Proposition 2 in our recent paper [31]" (published November 2023, near submission). While not per se improper, it is consistent with dividing one study into multiple mutually-citing units.

Evidence highlights

DOI: 10.1109/TSTE.2024.3388388
Contradiction location: Eq. (8c), Section II.D, vs. Eqs. (13)–(17), Section III.A.
Environment claim: Section IV, Numerical Simulation, opening paragraph ("MATLAB 2014a, Intel Core i5, 8 GB RAM").
Table III: participation factors on a coarse 0.1 grid across all three generators.
Citation chain: Reference [31], Theorem 1, Section III.A.

Notes

Image reuse (Western blot, microscopy, etc.) and Photoshop-level noise analysis were not applicable; the manuscript did not provide raw high-resolution figures suitable for pixel-level comparison.
Statistical p-hacking diagnostics were not applicable; no replicated experimental observations were supplied.
A 20-second wall-clock for the full MISOCP on a 118-bus hydrogen system with modern benchmarks is not impossible, but combined with the 10-year-old toolchain it is treated as implausible rather than impossible.
All numeric identifiers (DOIs, equation numbers, table numbers, k = 1,…,K) are reproduced verbatim from the source report.
This assessment is preliminary; final determination of academic misconduct requires investigation by the publisher or institutional body.

Suggested next actions

[x] Re-derive Eqs. (8c) → (13)–(17) to classify the mismatch as a notation slip or a structural error.
[ ] Request the authors clarify why a multivariate ambiguity set is paired with a scalar expectation reformulation.
[ ] Request original code and a reproducibility run on a current solver/hardware stack.
[ ] Consider a post-publication notice if the inconsistency is confirmed as fundamental rather than typographical.