Dynamic assessment of situation awareness in road tunnels: Considering tunnel light environment characteristics and drivers’ physiological perception states

Abstract

Interactions with varying light conditions in the road tunnels create a complex and challenging driving environment, leading to dynamic variations in drivers’ situation awareness, which are difficult to characterize. In cognizance of the unique light environment characteristics of tunnels, this study proposes a novel dynamic assessment method for situation awareness, aimed at accurately assessing drivers’ dynamic situation awareness level (DSAL) in tunnel sections yet without intruding on the driving task. First, a quantitative model of the visual perception ability is constructed based on the actual tunnel light environment information and the drivers’ visual perception characteristics. Additionally, an improved model for attention resource allocation, considering the stimulus of the tunnel light environment, is developed. Second, drivers’ physiological perception states in the tunnel sections are represented by fully accounting for the interaction between their visual perception abilities and attention resource allocation mechanisms. Subsequently, the drivers’ dynamic response processes in hazard identification during tunnel driving are quantitatively described by combining their physiological perception states with the adaptive control of thought-rational. On this basis, a highly interpretable DSAL assessment model is constructed using Bayesian conditional probability theory. Finally, the effectiveness and advancement of the assessment method are validated for a case study of real-vehicle driving in eight tunnels. The results indicate that the proposed assessment method achieves an average relative error of 7.90% with a standard deviation of 4.08%, considerably lower than those of other existing non-intrusive assessment methods. Therefore, the DSAL assessment results are closer to actual situation awareness and exhibit excellent stability, demonstrating strong potential for practical applications.