KOALA2 - cooperative interacting vehicles with low latency communication

无线通信系统在过去几年中有了很大的发展，可以利用先进的技术来提高车辆的安全性和交通效率。特别是，与传统通信系统相比，WLAN-V2X和蜂窝V2X显著缩短了通信延迟。虽然SPP1835第一阶段的考拉项目主要关注减少通信系统产生的延迟以及对延迟进行建模，但后续项目KOALA2解决了如何以整体方法将这些低延迟通信系统用于基于通信的车辆自动化。KOALA2有三个主要目标：(1)用于协作感知的车辆之间的传感器数据交换很容易超过通信链路的可用数据速率。与以前的数据拥塞控制方法不同，考虑到当前的驾驶情况，KOALA2研究了动态调整生成的对象列表以匹配瞬时数据速率的自适应方案。(Ii)对于协作用例中车辆之间的基于场景的机动协调(如换道、排队和紧急转向)，该项目开发了用于通信的设计模式，以通过显式通信来补充基于传感器的机动规划。最先进的基于广播的信息传播是通过以车辆组进行通信来扩展的，车辆组仅包括相应演习的参与者。设计模式被用在一个消息传递框架中，该消息传递框架实现了协同驾驶动作的规划、执行和完成(或终止)的可靠通信。(3)所开发的技术是典型实施和实验验证的。该项目可持续地改进协同感知以及协同机动和轨迹规划作为协同自动驾驶中与安全相关的时间关键用例的关键技术。

Wireless communication systems have considerably evolved over the last years and advanced technologies for improving vehicle safety and traffic efficiency are available. In particular WLAN-V2X and Cellular V2X considerably shorten, in comparison to conventional communication systems, the communication latency. While the project KOALA in the first phase of SPP1835 mainly focused on the reduction of the latency generated by the communication system as well as on modelling the latency, the follow-up project KOALA2 addresses how these low latency communication systems can be exploited for communication-based vehicle automation in a holistic approach. KOALA2 has three main objectives: (i) The exchange of sensor data among vehicles for cooperative perception can easily exceed the available data rate of the communication link. In contrast to previous approaches for data congestion control, KOALA2 studies adaptive schemes that dynamically adjust the generated object lists to match the instantaneous data rate, in consideration of the present driving situation. (ii) For scenario-based maneuver coordination among vehicles in cooperative use cases (such as lane change, platooning and emergency steering) the project develops design patterns for communication, in order to complement sensor-based maneuver planning by explicit communication. The state-of-the-art broadcast-based information dissemination is extended by communication in groups of vehicles, which exclusively encompasses the actors of the corresponding maneuver. The design patterns are used in a messaging framework that realizes a reliable communication for planning, execution and completion (or termination) of cooperative driving maneuvers. (iii) The developed techniques are prototypically implemented and experimentally validated. The project sustainably improves the cooperative perception as well as the cooperative maneuver and trajectory planning as key technologies for safety-relevant time critical use cases in cooperative automated driving.