Kim Inyoung, Kang Donghyo, Jeong Harim, Lee Soomok, Yun Ilsoo
Department of D.N.A. Convergence, Ajou University, Suwon 16499, Republic of Korea.
Department of Transportation System Engineering, Ajou University, Suwon 16499, Republic of Korea.
Sensors (Basel). 2023 Oct 6;23(19):8271. doi: 10.3390/s23198271.
With advances in the technology applied to automated driving systems (ADSs), active efforts have been made to evaluate the safety of ADS in various complex situations using simulations. In accordance with these efforts, numerous institutions have developed single-scenario pools that reflect a variety of road and traffic characteristics and ADS performances. However, a single scenario has limitations in comprehensively evaluating the performance of complex ADS. Therefore, this study proposed a methodology that combines and transforms single scenarios into multiple scenarios. This aided in continuously evaluating the ADS performance over entire road segments and implemented this methodology in the simulations.
随着应用于自动驾驶系统（ADS）的技术不断进步，人们积极努力通过模拟来评估ADS在各种复杂情况下的安全性。根据这些努力，众多机构开发了反映各种道路和交通特征以及ADS性能的单一场景库。然而，单一场景在全面评估复杂ADS的性能方面存在局限性。因此，本研究提出了一种将单一场景组合并转换为多个场景的方法。这有助于在整个路段上持续评估ADS性能，并在模拟中实施了该方法。
