A hierarchical hybrid control system is proposed to cope with highly automated driving in highway environments with multiple lanes and surrounding vehicles. In the high-level layer, the discrete driving decisions are coordinated by the finite-state machine (FSM) based on the relative position identification and predictive longitudinal distance of the surrounding vehicles. The low-level layer is responsible for the vehicle motion control, where the model predictive control (MPC) approach is utilized to integrate the longitudinal and lateral control mainly including car-following control and lane changing control. The proposed control system focuses on two issues regarding safe driving on highways. On one hand, the subject vehicle must always keep a safe distance with its leading vehicle to avoid the rear-end collision. On the other hand, the subject vehicle should also overtake the preceding vehicle by safe lane changes if the desired speed is not achieved. The effectiveness of the hybrid control is tested in the simulation, whose results verify that the driving decisions are made reasonably and the vehicle motion control obeys stability and comfort requirements. Moreover, it is also indicated by the simulations in random scenarios that the control strategy is able to deal with most of ordinary situations on highways although some emergency situations or critical driving maneuvers of other vehicles are not considered.

提出了一种分层混合控制系统，以应对在具有多车道和周围车辆的高速公路环境中的高度自动化驾驶。在高层中，离散驾驶决策由有限状态机（FSM）根据相对位置标识和周围车辆的预测纵向距离进行协调。低层负责车辆的运动控制，其中模型预测控制（MPC）方法用于整合纵向和横向控制，主要包括跟车控制和换道控制。拟议中的控制系统着重于与高速公路安全驾驶有关的两个问题。一方面，目标车辆必须始终与其前车保持安全距离，以免发生追尾事故。另一方面，如果未达到所需的速度，则本车辆还应通过安全车道变更超车。仿真中测试了混合动力控制的有效性，其结果证明了合理地做出驾驶决策，并且车辆运动控制符合稳定性和舒适性要求。此外，在随机情况下的仿真也表明，尽管未考虑某些紧急情况或其他车辆的关键驾驶操作，但控制策略能够应对高速公路上的大多数普通情况。其结果验证了合理的驾驶决策，并且车辆运动控制符合稳定性和舒适性要求。此外，在随机情况下的仿真也表明，尽管未考虑某些紧急情况或其他车辆的关键驾驶操作，但控制策略能够应对高速公路上的大多数普通情况。其结果验证了合理的驾驶决策，并且车辆运动控制符合稳定性和舒适性要求。此外，通过随机场景中的仿真还表明，尽管未考虑某些紧急情况或其他车辆的关键驾驶操作，但控制策略能够应对高速公路上的大多数普通情况。