The invention provides a ramp afflux strategy learning method and device, electronic equipment and a storage medium, and relates to the technical field of automatic driving. The strategy learning method comprises the following steps: configuring an accumulated expected return in a training environment, the accumulated expected return being used for representing an accumulated value of an expected return for continuously executing an afflux strategy to complete a ramp afflux process; a strategy evaluation model and a strategy improvement model are constructed based on the accumulated expected return, the strategy evaluation model is used for evaluating a first expected return obtained by following an afflux strategy from a specified driving state, and the strategy improvement model is used for evaluating a second expected return obtained by adopting a specified driving action; alternative iterative learning is carried out on the strategy evaluation model and the strategy improvement model, and distributed driving actions are obtained; and performing collision avoidance correction operation on the distributed driving action based on the safety shield model, and generating a corrected target afflux strategy. Through the technical scheme disclosed by the invention, the safety of the output strategy can be improved when the method is applied to a scene of converging from a ramp to an expressway.
本公开提供了一种匝道汇入策略学习方法、装置、电子设备和存储介质,涉及自动驾驶技术领域。其中,策略学习方法包括:在训练环境配置累计期望回报,累计期望回报用于表示连续执行汇入策略以完成匝道汇入过程的期望回报的累计值;基于累计期望回报构建策略评估模型与策略改进模型,策略评估模型用于评估从指定行驶状态起遵循汇入策略得到的第一期望回报,策略改进模型用于评估采用指定行驶动作得到的第二期望回报;对策略评估模型与策略改进模型进行交替迭代学习,得到分布行驶动作;基于安全护盾模型对分布行驶动作进行避撞修正操作,生成修正后的目标汇入策略。通过本公开的技术方案,能够在应用于从匝道汇入高速公路的场景时提高输出策略的安全性。
Ramp afflux strategy learning method and device, electronic equipment and storage medium
匝道汇入策略学习方法、装置、电子设备和存储介质
2022-07-29
Patent
Electronic Resource
Chinese
IPC: | G06F ELECTRIC DIGITAL DATA PROCESSING , Elektrische digitale Datenverarbeitung / G06N COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS , Rechnersysteme, basierend auf spezifischen Rechenmodellen / G08G Anlagen zur Steuerung, Regelung oder Überwachung des Verkehrs , TRAFFIC CONTROL SYSTEMS |
Vehicle ramp afflux method, device and equipment and storage medium
European Patent Office | 2023
|Cooperative vehicle ramp afflux method and device, computer equipment and storage medium
European Patent Office | 2023
|Vehicle ramp afflux-in and afflux-out cooperative control method
European Patent Office | 2022
|Intelligent motorcade collaborative afflux method and device and electronic equipment
European Patent Office | 2023
|