New NMPC framework lets drivers drift with AI-assisted stability
Driver stays in control while NMPC handles stability via steer‑by‑wire.
Deep Dive
A paper proposes a nonlinear MPC framework for assisted drifting on rear‑wheel‑drive vehicles. Using steer‑by‑wire and drive‑by‑wire interfaces, the controller decouples driver commands from direct actuator inputs so the driver regulates sideslip via the steering wheel while the NMPC maintains stability. A dedicated activation logic ensures engagement only under deliberate intent. High‑fidelity simulations show stabilization even with continuously varying sideslip references, using a simple single‑track model with basic tire dynamics.
Key Points
- Nonlinear MPC framework for assisted drifting keeps driver in control via steer‑by‑wire and drive‑by‑wire interfaces.
- Activation logic ensures controller only engages under deliberate driver intent, avoiding unwanted assistance.
- High‑fidelity simulations show stabilization with a simple single‑track model and basic tire dynamics, even with continuously varying sideslip references.
Why It Matters
Makes advanced drifting techniques safer and more accessible to non‑expert drivers, blending human skill with AI stability control.