Optimizing AUV speed dynamics with a data-driven Koopman operator approach

Zhiliang Liu, Xin Zhao, Peng Cai, Bing Cong

wrongmedium confidence

Category: math.DS
Journal tier: Specialist/Solid
Processed: Sep 28, 2025, 12:56 AM
arXiv Links: Abstract ↗PDF ↗

Audit review

The paper’s EDMD-MPC pipeline is plausible but contains a critical dimensional-error in the increment-augmented lifted dynamics (its equation (25) uses [A B; 0 B] and [0; B] where the correct blocks are [A B; 0 I] and [B; I]) and omits the ridge-EDMD normal equations and condensed QP derivations. The model solution supplies the correct ridge solution for [A B], the proper augmentation with Δu, and a complete convex QP condensation with a strictly positive-definite Hessian, yielding uniqueness when feasible.

Referee report (LaTeX)

\textbf{Recommendation:} major revisions

\textbf{Journal Tier:} specialist/solid

\textbf{Justification:}

The contribution—Koopman EDMD identification combined with MPC enforcing both input and increment constraints—is practical, but the manuscript has a serious error in the augmented dynamics (eq. (25)) and omits key derivations (ridge solution, condensed QP, convexity/uniqueness). Simulations suggest the approach is viable, yet the theoretical presentation must be corrected and completed to be publishable. With a corrected augmentation ([A B; 0 I], [B; I]), explicit ridge formulas, and a proper condensed QP exposition, the paper could meet standards for a solid specialist venue.