The invention discloses a path tracking stability control method based on reinforcement learning, and the method comprises the steps: 1, analyzing a wheel-ground interaction mechanism of an off-road vehicle, building a wheel-ground interaction pressure-bearing and shearing model, and calculating the force borne by the vehicle after the wheel-ground interaction is considered, on the basis, an inner rollover evaluation index and a sideslip evaluation index are constructed; and 2, the tracking stability under the complex terrain is controlled through a reinforcement learning-based TD3 algorithm, and a state space, an action space, a reward function and a network structure of the reinforcement learning algorithm are designed, so that an expected action is output through an active steering and longitudinal force optimal distribution mode, and a control effect is achieved. According to the invention, the path tracking of the off-road vehicle under the complex terrain can be realized, and the precision and stability in the tracking process are ensured.
本发明公开了一种基于强化学习的路径跟踪稳定性控制方法,包括:1分析越野车辆车轮‑地面相互作用机理,建立车轮‑地面相互作用承压和剪切模型,并计算出考虑车轮‑地面相互作用后车辆受到的力,以此为基础构建里侧翻评价指标和侧滑评价指标;2通过基于强化学习的TD3算法对复杂地形下的跟踪稳定性进行控制,并设计强化学习算法的状态空间、动作空间、奖励函数和网络结构,从而通过主动转向和纵向力优化分配的方式输出期望的动作,以达到控制效果。本发明能够实现越野车辆在复杂地形下的路径跟踪,保证跟踪过程中的精度和稳定性。
Path tracking stability control method based on reinforcement learning
一种基于强化学习的路径跟踪稳定性控制方法
2024-08-09
Patent
Electronic Resource
Chinese
IPC: | B60W CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION , Gemeinsame Steuerung oder Regelung von Fahrzeug-Unteraggregaten verschiedenen Typs oder verschiedener Funktion |
Reinforcement Learning-Based Robust Control for Path Tracking of Automated Vehicles
Springer Verlag | 2024
|Path Tracking Control of Hybrid-driven Robotic Fish Based on Deep Reinforcement Learning
British Library Conference Proceedings | 2020
|Deep reinforcement learning based path tracking controller for autonomous vehicle
SAGE Publications | 2021
|European Patent Office | 2024
|