High-Fidelity Co-Simulation-Based Validation of Advanced Truck Platooning Control
Road freight transportation demand will further increase in the coming years. This leads to challenges for the freight sector such as steadily increasing energy consumption and the coping with limited traffic system capacity. One answer to these challenges is cooperative vehicle platooning. Related state-of-the-art vehicle automation technologies can increase traffic throughput, improve road safety, and reduce fuel consumption. We developed a holistic distributed control concept to enable tight vehicle spacing and efficient platoon maneuvers with the option of utilizing vehicle-to-vehicle communication while providing collision safety at all times. This innovative platoon control architecture combines trajectory optimization and local model-predictive control of each vehicle.
The maneuver-specific reference trajectories for each platooning vehicle are optimized and communicated by the Platoon Coordinator. Each vehicle is locally controlled by model-predictive control (MPC) that incorporates the information received from the Platoon Coordinator. The MPC law is specifically formulated so that a safe stop in case of a sudden emergency braking maneuver of the preceding car is always possible. Also, communicated predictions and agreements among neighboring platooning vehicles are carefully considered to improve efficiency. The proposed concept remains highly scalable due to its distributed control structure. The local MPC is further enhanced by innovations to reduce safely realizable vehicle spacing and increase communication efficiency. It is essential to validate such complex control concepts in a realistic virtual environment. To do so, we have developed a suitable co-simulation platform: The proposed control concepts are validated via realistic co-simulations of high-fidelity vehicle dynamics for the exemplary case of truck platooning. Therefore, each individual vehicle is simulated by IPG TruckMaker®, while MATLAB® provides the simulation environment and Simulink® is used as communication interface between the individual vehicle instances. Additionally, a real-time-capable driving simulator was developed to allow human drivers to test the implemented control concepts in a human-in-the-loop setting.