Infinite-horizon model predictive control of path tracking control of autonomous vehicles subject to actuator delays

Autonomous vehicles are often subject to various actuator delays, which pose significant challenges for path-tracking controller design, especially for implementation on automotive chips with limited computational resources. To address this, this paper proposes an infinite-horizon model predictive control (IMPC) scheme. The IMPC is designed using dynamic programming, and integrates the linear quadratic regulator (LQR) with model predictive control (MPC). This approach reduces online computation to simple matrix operations and table lookups, making it suitable for resource-limited automotive chips. The proposed method guarantees the stability of the delayed closed-loop system and offers superior computational efficiency compared to conventional MPC. Theoretical analysis, simulations, and experimental results demonstrate the effectiveness and advantages of the proposed approach.