Abstract: In this study, four test scenarios are established to analyze drivers′ deceleration behavior at tunnel entrances during daytime free-flow conditions and quantitatively evaluate the regulating effects of different visual guiding facilities. The scenarios include: Scenario 1 (control group with industry standards), Scenario 2 (2.5 m elevation marker), Scenario 3 (circular elevation marker), and Scenario 4 (multi-level visual guiding system). The deceleration start position, changes in deceleration rate, and throttle opening rate are used as analytical indices. The test results show the following: (1) Drivers′ speed patterns generally follow a stable-declining to relatively stable trend. (2) In Scenario 1, drivers began decelerating later, which lead to a more abrupt deceleration and a shorter deceleration stage, resulting in less driving comfort and higher control risk compared to other scenarios. (3) Scenarios 2 and 3 allow drivers to identify the tunnel portal earlier, thereby extending the driving comfort zone. (4) Scenario 4 performs the best, with the deceleration starting position 65% ahead of that in Scenario 1, and the driving comfort zone extends by 13.1% compared to Scenario 1. Furthermore, Scenario 4 achieves the highest average throttle opening (0.095) and the lowest throttle control risk (0.13), making it the most suitable deceleration mode for tunnel entrance areas. The test results demonstrate that using circular elevation markers at tunnel and a multi-level visual guiding system at tunnel entrances helps drivers detect the entrance earlier, start decelerating sooner, and decelerate slowly. This significantly improves driving safety and comfort in tunnel entrance areas.

Key words: expressway tunnel, traffic safety, tunnel entrance zone, deceleration behavior, visual guiding system