Impact of speed limit strategies on the coupling between drivers' approaching and lane-changing behavior in work zones considering human heterogeneity.

OBJECTIVES: Speeding while approaching highway work zones reduces the time available for drivers to identify hazards, adjust speed, and prepare for lane changes. However, the extent to which alternative speed-limit configurations affect approach and lane-changing behaviors, the coordination between these behavioral domains, and response heterogeneity across driver groups remains insufficiently quantified. This study compared driver responses across five speed-limit schemes.

METHODS: A driving-simulator experiment was conducted to collect behavioral data from 37 licensed drivers. Approach behavior was characterized by average approach speed and maximum deceleration, whereas lane-changing behavior was characterized by lane-change decision distance and minimum conflict distance. A coupling coordination degree model was used to quantify the relative coordination between the two behavioral domains. Generalized estimating equations (GEE) were used to examine the main effects of speed-limit scheme, gender, and professional driving status, as well as the interactions of speed-limit scheme with gender and professional driving status, on the four behavioral measures.

RESULTS: Average approach speed, maximum deceleration, lane-change decision distance, and minimum conflict distance differed significantly across speed-limit schemes, with the relative rankings of the five schemes varying among the four behavioral measures. The 300-m graded scheme was associated with a lower approach speed, earlier lane-change initiation, and a larger post-lane-change spatial margin and yielded the highest coupling coordination degree (0.995). The GEE models identified significant speed-limit scheme × gender and speed-limit scheme × professional driving status interactions for all four behavioral measures. Female drivers generally approached at higher speeds than male drivers but initiated lane changes earlier and maintained larger post-lane-change spatial margins. The effects of professional driving status varied across speed-limit schemes; under the non-graded schemes, drivers without professional driving experience exhibited stronger braking responses and later lane-change initiation.

CONCLUSIONS: The proposed framework integrates the relative coordination between approach and lane-changing behaviors with heterogeneity in driver responses, providing a quantitative reference for the comparative evaluation of work-zone speed-limit schemes and the subsequent design and validation of behavior-responsive speed-management measures.