New framework lets autonomous cars steer accurately while accelerating and braking
A robust nonlinear control that adapts to changing speed and vehicle uncertainty in real time.
A new robust nonlinear lateral control framework for autonomous vehicles accounts for varying longitudinal speed and acceleration. It uses feedback linearization to embed these dynamics into the control law and analyzes internal stability. Two robust designs are proposed: a Lyapunov redesign inspired by sliding mode control and an incremental nonlinear dynamic inversion (INDI) method. Simulations show enhanced tracking accuracy and robustness to parameter uncertainty, while real-vehicle tests confirm real-time path-tracking on actual hardware.
- Control law adapts to varying longitudinal speed (5-30 m/s) and acceleration, unlike existing constant-speed assumptions.
- Two robust designs (Lyapunov redesign and INDI) handle parameter uncertainties without needing exact vehicle models.
- Real-vehicle tests confirm practical path-tracking performance on actual hardware, not just simulation.
Why It Matters
Enables autonomous vehicles to steer safely during stop-and-go traffic and uncertain road conditions, improving real-world reliability.