Self-regulation during the takeover in automated driving refers to the driver’s proactive and deliberate behavioral adjustments aimed at ensuring safety and efficiency when responding to takeover scenarios in highway tunnels. However, prior research has not yet thoroughly explored the underlying mechanisms influencing such behavior. This study, grounded in the theory of planned behavior and employing structural equation modeling, investigates the mechanisms affecting drivers’ self-regulating behavior in highway tunnel takeover scenarios. It incorporates extended factors such as drivers’ degree of familiarity, risk perception, and exposure to risky situations. The study gathered 306 valid samples through an online survey method. The model results reveal that the intention to self-regulate is the most significant factor influencing self-regulating behavior, with a standardized path coefficient of 0.846. Secondary factors include attitudes toward self-regulation, perceived behavioral control, and subjective norm, which have standardized path coefficients of 0.460, 0.292, and 0.274, respectively. The results of the mediation analysis indicate that self-regulating intention and degree of familiarity serve as mediating factors. In addition, the XGBoost-SHAP model was used to evaluate the overall contribution of each variable to self-regulating behavior, and the findings are highly consistent with those obtained from the structural equation modeling.
Zhu et al. (Mon,) studied this question.