This study analyzes the impact of different numbers of visible retroreflective rings on the drivers’ ability to perceive the curvature in curved freeway tunnels under different small radii. A driving simulator was used to conduct simulation tests on the two visual guiding schemes in curved tunnels with chevron alignment signs and retroreflective rings. The drivers’ curve illusion and reaction time were tested to assess their ability to perceive curvature; the tests data were analyzed using OriginLab. In the unimproved scenes, the drivers overestimated the radii, at least, 35%, and the values of reaction time were at most 5.251 s. In the scenes where four visible chevron alignment signs were used, the curve illusion and reaction time under different radii were marginally decreased; only at the radii of 200 m (6.03%) and 400 m (6.12%), the drivers could reach a better perception of curvature. In the scenes with three visible retroreflective rings, the degree of curve illusion under different radii was at most 5%, and the values of reaction time were, at least, 37% lower than the unimproved conditions. The scheme of three visible retroreflective rings was more conducive to improving the drivers’ perception of curvature and controlling the reaction time within a reasonable range. The fitting analysis of the logistic function revealed that the fitted curves’ inflection points between curve illusion and the number of retroreflective rings under different radii were close to 3, where the degree of the curve illusion reached the lowest. Meanwhile, the drivers’ reaction time under different radii negatively correlated with the number of visible retroreflective rings. When the number of retroreflective rings was more than four, the reaction time exhibited a convergence trend, suggesting that the increasing number of visible retroreflective rings exerted a limited effect on shortening the reaction time. Hence, this study recommends utilizing three visible retroreflective rings to guide the drivers’ sight in the curved freeways tunnels.

这项研究分析了不同数量的可见逆反射环对驾驶员感知不同小半径下的弯曲高速公路隧道曲率的能力的影响。在具有人字形对准标志和逆反射环的弯曲隧道中，使用驾驶模拟器对两种视觉引导方案进行了模拟测试。测试了驾驶员的曲线错觉和反应时间，以评估驾驶员感知曲率的能力；使用OriginLab分析测试数据。在未经改善的场景中，驾驶员至少高估了半径35％，反应时间最多为5.251 s。在使用四个可见人字形对准标志的场景中，不同半径下的曲线错觉和反应时间略有减少；仅在半径为200 m（6.03％）和400 m（6.12％）时，驾驶员可以更好地感知曲率。在具有三个可见的逆反射环的场景中，不同半径下的曲线幻觉度最多为5％，并且反应时间的值至少比未改善的条件低37％。三个可见的逆反射环的方案更有利于改善驾驶员的曲率感知并将反应时间控制在合理范围内。对数函数的拟合分析表明，在不同半径下，曲线幻觉与逆向反射环数之间的拟合曲线拐点接近3，曲线幻觉度达到最低。同时，驾驶员在不同半径下的反应时间与可见反光环的数量呈负相关。当逆反射环的数目大于四个时，反应时间呈现收敛趋势，这表明可见逆反射环数目的增加对缩短反应时间的作用有限。因此，这项研究建议利用三个可见的逆反射环来引导驾驶员在弯曲的高速公路隧道中的视线。