Answers are grounded in this paper and cite their sources.
Reliable three-dimensional trajectory tracking of autonomous underwater vehicles under ocean currents, turbulence, and actuator constraints challenges existing control frameworks. This paper proposes RUN-ESFM, a chaotic enhancement of the Runge–Kutta optimizer built on a novel one-dimensional Exponential Sinusoidal Feedback Map (1D-ESFM), defined by f ( x ) = sin π ⋅ mod r ⋅ x ⋅ e λ x ( 1 − x ) , 1 , exhibiting formal local expansivity and strong statistical randomness validated across all 15 NIST SP800-22 tests. Four targeted injection points integrate the ESFM sequence into RUN, complemented by adaptive cosine annealing and golden-ratio seeding. Validated on CEC2022 (dimension 100, 30 runs) against six competing algorithms, including four recent metaheuristics, RUN-ESFM achieves the best mean fitness on 6/12 functions and the lowest average rank of 1.58. The algorithm is deployed as a real-time MPC controller for BlueROV2 3D path tracking, a 12-dimensional problem solved within 1000 evaluations per step over a 120 s helical trajectory under offshore disturbances (current, JONSWAP waves, Von Kármán turbulence). Benchmarked against ten competing solvers, including a numerical baseline, and averaged over 15 independent disturbance realizations, RUN-ESFM achieves the lowest mean RMSE (0.4368 m), lowest mean control effort (3026.78 N ⋅ s), and fastest mean convergence (2.5 s), ranking first with a composite score of 3.25.
No takes yet. Share an insight, caveat, or question.
Tahiri et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: