Digital Twin-Based Trajectory Planning and Feedforward-Feedback Control for an Autonomous Vehicle on Urban Roads

In this paper, trajectory planning is realized for an autonomous vehicle in urban road environments based on a digital twin (DT) system. A feedforward-feedback control framework is also proposed to complete virtual-physical synchronous in the DT system. An A ${}^{\ast }$ algorithm is improved to obtain a global path based on a high-precision map in OpenDRIVE format. The global path is utilized by a virtual vehicle with a model predictive controller to produce a reference trajectory and feedforward inputs for a physical vehicle. A backstepping controller with the feedforward inputs is designed for the physical vehicle to track the reference trajectory. Experimental results demonstrate effectiveness of the trajectory planning and feedforward-feedback control method based on the DT system.