Predictive Iterative Learning Control for High-Precision Longitudinal Coordination of Vehicular Platoons

This paper addresses the longitudinal control problem of autonomous vehicular platoons, with the objective of achieving high-precision coordination through predictive iterative learning control (PILC). The proposed PILC method integrates past coordination experience and future predictions of the following vehicle to enhance control accuracy and accelerate learning convergence. Specifically, a super-lifted model is developed, and the convergence condition of the linear time-varying system is derived. The longitudinal platoon control problem is tackled by designing a PILC-based feedforward-feedback coordination controller. Simulation results demonstrate the fast convergence and robustness of PILC, while real-world experiments on a vehicular platoon platform validate its superior precision compared with baseline platoon controllers.