Behavior planning for autonomous vehicles

Embodiments of the present disclosure relate to behavioral planning for autonomous vehicles. Techniques described herein select a preferred trajectory for an autonomous vehicle based on evaluation of a plurality of hypothetical trajectories by different components within a planning system. Each component provides an optimization score for each trajectory according to the priority of the component, and scores from multiple components may form a final optimization score. The scoring system allows contemporaneous consideration of competitive priorities (e.g., comfort, shortest travel time, fuel economy) for different components. In an example, a trajectory with an optimal combination score may be selected for implementation. Thus, the optimal or preferred trajectory of the autonomous vehicle can be identified while navigating the environment using an iterative approach that evaluates various factors. 本公开的实施例涉及自主车辆的行为规划。本文所述技术基于规划系统内不同组件对多个假设轨迹的评估来选择自主车辆的优选轨迹。各个组件根据组件的优先级为每个轨迹提供优化分数，并且来自多个组件的分数可以形成最终的优化分数。该评分系统允许