While highly automated driving relies most of the time on a smooth driving assumption, the possibility of a vehicle performing harsh maneuvers with high dynamic driving to face unexpected events is very likely. The modeling of the behavior of the vehicle in these events is crucial to proper planning and controlling; the used model should present accurate and computationally efficient properties to ensure consistency with the dynamics of the vehicle and to be employed in real-time systems. In this article, we propose an LSTM-based hybrid extended bicycle model able to present an accurate description of the state of the vehicle for both normal and aggressive situations. The introduced model is used in a Model Predictive Path Integral (MPPI) plan and control framework for performing trajectories in high-dynamic scenarios. The proposed model and framework prove their ability to plan feasible trajectories ensuring an accurate vehicle behavior even at the limits of handling.

翻译：尽管高度自动驾驶在大多数情况下依赖于平稳驾驶假设，但车辆为应对突发状况而执行高动态激烈操纵的可能性极高。在这些事件中对车辆行为的建模对于合理规划与控制至关重要；所用模型应具备精确且计算高效的特点，以确保与车辆动力学的一致性，并能应用于实时系统。本文提出一种基于LSTM的混合扩展自行车模型，该模型能够准确描述车辆在常规与激烈工况下的状态。所提出的模型被应用于模型预测路径积分（MPPI）规划与控制框架中，以执行高动态场景下的轨迹规划。实验证明，该模型与框架能够规划出可行的轨迹，即使在操控极限条件下也能确保精确的车辆行为。