The invention relates to an automatic driving vehicle control method based on multi-agent reinforcement learning, and belongs to the technical field of artificial intelligence and automatic driving. For an automatic driving vehicle scene, the method provides a multi-agent adaptive decision framework for AVs. Firstly, individual requirements of passengers are considered, and different driving styles are formulated for AVs for use in a subsequent decision modeling process. A multi-agent system is divided by using a coalition game and V2V communication and taking each networked autonomous vehicle (CAV) as a center, so that a cooperative decision-making area is constructed. Then, based on a rolling time domain in model prediction control, a deep reinforcement learning decision framework is constructed to realize motion decision of the CAV in the cooperation area. In a fully automatic vehicle driving scene, the safety and efficiency on a high-speed trunk lane can be ensured, and the safety, comfort and efficiency of a ramp confluence area can also be ensured.
本发明涉及一种基于多智能体强化学习的自动驾驶车辆控制方法,属于人工智能和自动驾驶技术领域。针对自动驾驶车辆场景,本方法提出了一个用于AVs的多智能体自适应决策框架。首先考虑乘客的个性化需求,为AVs制定不同的驾驶风格,用于后续决策建模过程中。利用联盟博弈和V2V通信,以每个网联自动驾驶车辆CAV为中心划分多智能体系统,从而构建合作决策区域。然后,基于模型预测控制中的滚动时域,构建了一个深度强化学习决策框架,以实现CAV在合作区域里的运动决策。在完全自动驾驶车辆的场景下,本发明不仅能够保证在高速主干车道上的安全性和效率,而且对于匝道合流区域的安全性、舒适度和效率也能够得到保障。
Automatic driving vehicle control method based on multi-agent reinforcement learning
一种基于多智能体强化学习的自动驾驶车辆控制方法
2023-07-07
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 / G06N COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS , Rechnersysteme, basierend auf spezifischen Rechenmodellen |
Automatic driving key scene generation method based on multi-agent reinforcement learning
European Patent Office | 2024
|Automatic driving vehicle interaction control method based on reinforcement learning
European Patent Office | 2024
|Hybrid electric vehicle ecological driving strategy based on multi-agent reinforcement learning
European Patent Office | 2025
|Automatic driving vehicle microscopic decision-making method based on reinforcement learning
European Patent Office | 2020
|