Adaptive cruise control method and system based on deep reinforcement learning
Adaptive cruise control method and system based on deep reinforcement learning
- New search for: ZHENG XUELIAN
- New search for: HAN ZHUOCHENG
- New search for: REN YUANYUAN
- New search for: LI XIANSHENG
- New search for: XI JIANFENG
- New search for: WANG JIE
- New search for: LI YUANZHAO
- New search for: WU XUEFENG
- New search for: ZHENG XUELIAN
- New search for: HAN ZHUOCHENG
- New search for: REN YUANYUAN
- New search for: LI XIANSHENG
- New search for: XI JIANFENG
- New search for: WANG JIE
- New search for: LI YUANZHAO
- New search for: WU XUEFENG
The invention provides an adaptive cruise control method and system based on deep reinforcement learning. A driver car-following style identification module established based on a BP neural network; building an ACC control model based on deep reinforcement learning; the ACC control model comprises a strategy module and a deep Q network module; dividing the vehicle following behaviors of the driver into three types of vehicle following styles and then labeling; the car-following style identification model is used for training by using a BP neural network to obtain a driver car-following style classification model, and outputting a car-following style of a driver based on the driver car-following style classification model; the strategy module establishes a Q value network, and outputs an expected acceleration for adjusting the vehicle through the Q value network according to the obtained effective characteristic performance data of the real-time vehicle following behavior; the self-adaptive cruise control method closer to the individuation of a driver is achieved, and the safety, the following performance, the comfort and the economical efficiency are high.
本发明提供一种基于深度强化学习的自适应巡航控制方法以及系统；基于BP神经网络建立的驾驶人跟车风格辨识模块；基于深度强化学习建立的ACC控制模型；ACC控制模型包括策略模块、深度Q网络模块；将驾驶人跟车行为分为三类跟车风格然后打标签；跟车风格辨识模利用BP神经网络训练得到驾驶人跟车风格分类模型，基于驾驶人跟车风格分类模型输出驾驶人的跟车风格；所述策略模块建立Q值网络，通过Q值网络根据获取的实时跟车行为的有效特征性能数据，输出调整车辆的期望加速度；实现更接近驾驶人员个性化的自适应巡航控制方法且安全性、跟随性、舒适性、经济性高。
Document information
Adaptive cruise control method and system based on deep reinforcement learning
一种基于深度强化学习的自适应巡航控制方法以及系统
2023-06-13
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 |
Similar titles
Multi-Objective Adaptive Cruise Control via Deep Reinforcement Learning
| SAE Technical Papers | 2022
Multi-Objective Adaptive Cruise Control via Deep Reinforcement Learning
| British Library Conference Proceedings | 2022
Multi-Objective Adaptive Cruise Control via Deep Reinforcement Learning
| British Library Conference Proceedings | 2022
Multi-Objective Adaptive Cruise Control via Deep Reinforcement Learning
| British Library Conference Proceedings | 2022
Multi-Objective Adaptive Cruise Control via Deep Reinforcement Learning
| British Library Conference Proceedings | 2022