The invention relates to an automatic driving lane changing decision-making control method based on rule fusion reinforcement learning. The method comprises the following steps: establishing a longitudinal motion control model and a transverse behavior decision-making model of other vehicles; configuring a highway driving environment, screening out unreasonable traffic flow scenes, and constructing an automatic driving decision training scene; establishing a state space and an action space of driving behavior observation; building a deep learning network for strategy updating; designing a lane change decision risk assessment method, and establishing a decision safety evaluation mechanism; setting a reward function; introducing a security evaluation mechanism into a deep learning network based on DDQN; and performing fusion training based on an automatic driving decision training scene, correcting a decision-making dangerous action, setting two storage experience pools, updating parameters of a target value network in the model in combination with a reward function, and applying the trained model to automatic driving lane changing decision making. Compared with the prior art, the method has higher decision task success rate, driving efficiency and safety in a lane changing scene.
本发明涉及一种基于规则融合强化学习的自动驾驶换道决策控制方法,包括:建立他车纵向运动控制模型及横向行为决策模型;配置高速公路驾驶环境,筛除不合理车流场景,构建自动驾驶决策训练场景;建立驾驶行为观测的状态空间、动作空间;搭建用于策略更新的深度学习网络;设计换道决策风险评估方法,建立决策的安全性评判机制;设置奖励函数;将安全性评判机制引入基于DDQN的深度学习网络;基于自动驾驶决策训练场景进行融合训练,修正决策出的危险动作,并设置两个存储经验池,结合奖励函数更新模型中目标价值网络的参数,训练后的模型用于自动驾驶换道决策。与现有技术相比,本发明在换道场景中具有更高的决策任务成功率、行驶效率以及安全性。
Automatic driving lane changing decision control method based on rule fusion reinforcement learning
一种基于规则融合强化学习的自动驾驶换道决策控制方法
2022-11-01
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 / G06F ELECTRIC DIGITAL DATA PROCESSING , Elektrische digitale Datenverarbeitung / G06N COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS , Rechnersysteme, basierend auf spezifischen Rechenmodellen |
European Patent Office | 2024
|European Patent Office | 2022
|Automatic driving decision-making method based on rule-assisted reinforcement learning
European Patent Office | 2023
|Automatic driving automobile lane changing decision control method considering uncertainty
European Patent Office | 2022
|Automatic driving vehicle lane changing method based on double-layer deep reinforcement learning
European Patent Office | 2023
|