Answers are grounded in this paper and cite their sources.
Four-wheel independently driven electric vehicles can generate corrective yaw moments, but authority varies with traction demand, load transfer, combined slip, motor saturation, and rate limits. We propose allocation-aware prescribed-performance barrier direct yaw control (APB-DYC), which evaluates estimated wheel-level feasibility before command generation. Total torque is projected onto the wheel-torque box; motor and residual-friction limits then define an estimated attainable yaw-moment interval. Authority contraction and realization error adapt tracking envelopes, while a sampled-data barrier projects the nominal command onto attainable, rate-admissible, and handling-safe intervals. A strictly convex bounded equality program allocates the accepted demands by enumerating at most 81 wheel-status patterns. Interval exactness and finite allocator termination hold for the stated box model; barrier invariance and transformed-coordinate boundedness require bounded implementation error and a nonempty interval. A nonlinear seven-state numerical study covers four principal cases, an 18-point speed–adhesion grid, ablations, 14 independent perturbations, and added estimator, combined-error, friction-transition, missed-update, and plant-dynamics tests. APB-DYC produced the smallest combined tracking error among three method-level baselines. Seventeen grid cases retained a positive barrier minimum; one reached −0.094. All 12 added high-speed fishhook cases retained positive barrier minima, including controller-side lateral-force reconstruction, simultaneous friction-estimate/sideslip/lag perturbations, a 0.8-to-0.5 friction transition with 250 ms estimate lag, one held update, and augmented plant dynamics. The results support constrained yaw tracking inside the tested conditional operating domain.
No takes yet. Share an insight, caveat, or question.
Tao et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: