Real-Time Solution-Seeking for Game-Theoretic Autonomous Driving via Time-Distributed Iterations
The work introduces time-distributed Newton and Newton–Kantorovich methods to solve game-theoretic model predictive control (GT-MPC) for autonomous driving. By distributing solution-seeking iterations over time, the system can find real-time Nash equilibria in complex multi-agent scenarios. This approach significantly reduces computational complexity while ensuring fast, practical decision-making.