Optimal investment in driving automation: Individual vs. cooperative sensing
Optimal investment in driving automation: Individual vs. cooperative sensing
复制标题
驾驶自动化的最佳投资：个体传感与协作传感
DOI：
10.1016/j.trb.2023.06.001
复制
发表时间：
2023
期刊：
影响因子：
--
通讯作者：
Yin, Yafeng
中科院分区：
文献类型：
--
作者：
Nourinejad, Mehdi;Bahrami, Sina;Yin, Yafeng
联网自动车辆 (CAV) 使用传感器扫描周围环境，以便做出安全高效的运动决策。 CAV 视觉的丰富程度取决于车载传感器的配置以及 CAV 与附近车辆共享数据的能力。高端传感器提供高质量的数据，使 CAV 能够通过单独的传感来依赖自己的传感器，并独立于其他车辆。相比之下，传感器数据共享通过协作感知增强了 CAV 的集体视野。本研究调查了个人感知与合作感知的价值和成本的权衡，并提出了短期和长期规划的最佳投资策略。利用 CAV 专用走廊的设置，我们提出了两个非线性方案，以确定长期走廊容量的最佳投资，并通过道路定价实现短期社会福利最大化。容量分析显示，个人传感和协作传感的投资脱节，首先是前者，直到传感器达到所需的分辨率，然后是后者。社会福利分析显示，旅行者在某些环境下会获得奖励，鼓励实现合作感知的流动。在无法从协作传感中受益的人类驾驶车辆环境中，不会观察到这种奖励。
Connected automated vehicles (CAVs) use sensors to scan their surrounding environment in order to make safe and efficient motion decisions. The richness of a CAV’s vision depends on the configuration of on-board sensors and the ability of CAVs to share data with nearby vehicles. High-end sensors provide quality data, allowing CAVs to rely on their own sensors through individual sensing, and to become independent of other vehicles. In contrast, sensor data sharing enhances the collective vision of CAVs through cooperative sensing. This study investigates the trade-offs in the values and costs of individual versus cooperative sensing, and it proposes optimal investment strategies for short- and long-term planning. Exploiting the setting of a CAV-exclusive corridor, we propose two nonlinear programs to determine optimal investment in corridor capacity in the long-term, and to maximize social welfare in the short-term through road pricing. Capacity analysis shows disjointed investments in individual and cooperative sensing, first in the former until sensors reach a desired resolution, and then in the latter. Social welfare analysis shows travelers are granted a reward in certain settings, encouraging flows that achieve cooperative sensing. Such rewards are not observed in human-driven vehicle settings that cannot benefit from cooperative sensing.