Traffic-Rule-Aware Reachability Analysis for Motion Planning of Automated Vehicles

自动车辆运动规划的交通规则感知可达性分析

• 批准号: 513192618
• 负责人: Professor Dr.-Ing. Matthias Althoff
• 金额: --
• 依托单位: Lehrstuhl für Informatik VI: Robotik, Künstliche Intelligenz und Echtzeitsysteme
• 依托单位国家: 德国
• 项目类别: Research Grants
• 资助国家: 德国
• 项目状态: 未结题
• 来源: https://gepris.dfg.de/gepris/projekt/513192618?language=en
• 关键词: Traffic; Rule; Aware; Reachability; Analysis

Despite the importance of respecting traffic rules, most previous works on motion planning for autonomous vehicles ignore traffic rules or are not able to consider most of them. One of the reasons is that most traffic rules cannot be directly formulated as constraints for a motion planner. Exceptions exist, such as keeping the speed limit, but complicated rules require the formalization as temporal logic formulas. So far, there is no approach that can exactly or approximately rewrite temporal logic formulas as constraints so that they can be directly used in optimization-based motion planning algorithms. Our proposal addresses exactly this open research problem. We will develop novel methods to rewrite temporal logic formulas so that they can be efficiently used in motion planners. Furthermore, we will develop for the first time methods to compute reachable sets that consider traffic rules. This will enable us to develop essentially new methods for the co-design of reachability analysis and motion planning considering traffic rules. We expect that with our novel approach, traffic-rule-compliant motion plans will be obtained the faster, the more critical the situation is. In addition, we will automatically ignore less important traffic rules in order to ensure the compliance with more important rules in critical situations. We will demonstrate our approach not only in simulation, but also with our new research vehicle EDGAR, which has been partially financed by a DFG Major Research Instrumentation grant .

尽管尊重交通规则很重要，但以往大多数自动驾驶车辆的运动规划工作都忽略了交通规则，或者无法考虑其中的大部分。其中一个原因是，大多数交通规则不能直接作为运动规划者的约束来制定。存在例外，例如保持速度限制，但复杂的规则需要将其形式化为时态逻辑公式。到目前为止，还没有一种方法可以准确地或近似地将时间逻辑公式重写为约束，以便直接用于基于优化的运动规划算法。我们的建议恰恰解决了这一开放的研究问题。我们将开发新的方法来重写时间逻辑公式，以便它们可以有效地用于运动规划器。此外，我们将首次开发计算考虑交通规则的可达集合的方法。这将使我们能够开发出本质上新的方法，用于考虑交通规则的可达性分析和运动规划的联合设计。我们预计，使用我们的新方法，可以越快、越危急地获得符合交通规则的运动计划。此外，我们将自动忽略不太重要的交通规则，以确保在关键情况下遵守更重要的规则。我们将不仅在模拟中展示我们的方法，而且还将用我们的新研究工具EDGAR来演示我们的方法，它的部分资金来自DFG主要研究仪器拨款。