The invention provides a vehicle following speed control method based on deep reinforcement learning, and the method comprises the steps: extracting an automatic driving following pair in a guiding vehicle-automatic driving vehicle-following vehicle mode based on a real driving data set; a Markov reward process is designed; designing a reinforcement learning reward function based on the requirements of vehicle safety and comfort in the three-vehicle mode; constructing a vehicle following speed control network framework based on deep reinforcement learning; training the obtained network to obtain a vehicle following speed control model based on deep reinforcement learning; and testing the obtained network model, and verifying the safety and comfort of the model. According to the method, the contradiction between trail-and-error of traditional deep reinforcement learning and automatic driving safety is relieved.
本发明提供了基于深度强化学习的车辆跟驰速度控制方法,包括:基于真实驾驶数据集,提取“引导车—自动驾驶车辆—跟驰车”三车模式下的自动驾驶跟驰对;设计马尔可夫奖励过程;基于三车模式下车辆安全性及舒适度的要求,设计强化学习奖励函数;构建基于深度强化学习的车辆跟驰速度控制网络框架;对所得到的网络进行训练,得到基于深度强化学习的车辆跟驰速度控制模型;对所得到的网络模型进行测试,验证模型的安全性及舒适度。本发明缓解了传统深度强化学习“试错”与自动驾驶安全性之间的矛盾。
Vehicle following speed control method based on deep reinforcement learning
基于深度强化学习的车辆跟驰速度控制方法
2023-09-29
Patent
Elektronische Ressource
Chinesisch
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 |
HEV energy management method based on deep reinforcement learning in vehicle following environment
Europäisches Patentamt | 2020
|DOAJ | 2022
|HUMAN-LIKE AUTONOMOUS VEHICLE SPEED CONTROL BY DEEP REINFORCEMENT LEARNING WITH DOUBLE Q-LEARNING
British Library Conference Proceedings | 2018
|