SPP 1835: Kooperativ interagierende Automobile

这个优先项目聚集了一个跨学科的科学家小组，为涉及自动驾驶汽车的合作交通开发一个系统理论框架。SPP的每个项目都研究道路使用者之间的合作方法，包括至少一辆自动驾驶汽车。a)协同感知：对于信息检索，协同交互汽车不仅考虑车载传感器信息，还分析来自其他汽车传感器的通信信息。为了在这种“远程信息感知视界”中充分利用信息，研究了考虑延迟、时空不确定性和可能的循环信息流导致的信息个体质量的鲁棒融合方法。b)情景预测：基于对当前交通状况的感知，预测其他交通参与者的未来行为和轨迹。本主题涉及预测人员和其他车辆移动的方法，以及确保他人可以预测自己行为的方法。c)协同机动和轨迹规划：基于感知和预测的交通场景，研究协同轨迹的规划策略。规划过程涉及就不同的行动方针进行含蓄或明确的谈判，目的是优化一个全面和共同的质量标准。d)数据和信息库：协同交互的汽车将它们的知识聚集在一个集体数据和信息库中，并将这些信息提供给交通社区。SPP主要寻求获取可能直接影响战术行为决策的信息（例如，对当地交通规则和条件的语义信息的推断），而传统SLAM方法主要获取几何信息。e)系统人机工程学：为了使科学进步造福于人类，研究驾驶员和乘客与自动驾驶车辆之间的相互作用是很重要的，这样人类就可以接受自动驾驶行为，并在必要时对其进行修改。f)汽车协同互动的跨领域问题：优先项目应为该领域提供整体贡献。其中包括信息质量、认知性能和轨迹安全的度量。

This priority programme gathers an interdisciplinary group of scientists to develop a system-theoretical framework for cooperative traffic involving automated automobiles. Each project of the SPP investigates methods for cooperation between road users including at least one automated vehicle.The Priority Programme focuses on the following topics:a) Cooperative Perception: For information retrieval cooperatively interacting automobiles not only consider the sensor information onboard the vehicle but also analyze communicated information from sensors of other automobiles. For full use of the information in this "telematic horizon of perception" robust fusion methods are investigated that consider the individual quality of the information due to latencies, spatio-temporal uncertainties, and possibly cyclic information flows. b) Situation Prediction: Based on the perception of the current traffic situation the future behavior and trajectories of other traffic participants shall be predicted. This topic involves methods to anticipate the movement of people and other vehicles as well as methods ensuring that others may anticipate one’s own behavior. c) Cooperative Maneuver and Trajectory Planning: Based on the perceived and predicted traffic scenes planning strategies for cooperative trajectories shall be investigated. The planning process involves implicit or explicit negotiations on different courses of action, aimed at optimizing a holistic and common quality criterion. d) Data and Information Base: Cooperatively interacting automobiles aggregate their knowledge in a collective data and information base and provide this information to the traffic community. The SPP primarily seeks to acquire information that may directly influence tactical behavior decisions (eg, inference of semantic information about local traffic rules and conditions) as contrasted to mainly geometric information acquired by traditional SLAM approaches. e) System Ergonomics: In order to make scientific progress for the benefit of humans, studies on the interaction between drivers and passengers with automated vehicles are important so that humans accept automated behavior and modify it when necessary. f) Cross-cutting Issues of Cooperatively Interacting Automobiles: The priority programme shall provide holistic contributions to the field. The include the imposture of metrics for information quality, cognitive performance, and trajectory safety.