Accounting about 70% of vehicle miles on roadways, highway driving is a critical issue in traffic safety deployment. Of the various maneuvers that comprise the highly complex driving task, each one requires understanding on the connections between driving states, vehicle performance and drivers' actions. This paper attempts to flesh out a complete picture of driving maneuvers on highways. Eighteen drivers participated in this study. They drove an instrumented vehicle on highways to accumulate 2,600 km naturalistic driving data. The data were segmented and classified into 11 maneuver groups manually. Analysis on the maneuvers revealed that: 1) A maneuver transition probabilities model was proposed. According to this model, 7 typical driving patterns were drawn based on the transition probabilities. Transition events pertaining to approaching/following/lane changing accounted for 95% of all the highway transition events. 2) The durations were 7.6/6.6 s and 7.1/7.0 s for free left/right lane changes and overtake from left/right lane changes, respectively. The numbers were 22.5, 21.4 and 16.3 s for far, middle and near following maneuvers, respectively. Statistical significances were found within both groups. 3) How drivers behave in each maneuver was analyzed. Drivers drove faster in free lane changes than did in overtake lane changes. For overtake lane changes, two driving patterns were observed: accelerate to change lane and decelerate to change lane.
Driving Maneuvers Analysis Using Naturalistic Highway Driving Data
2015-09-01
571667 byte
Conference paper
Electronic Resource
English
Deep Learning Approach for Detecting Lane Change Maneuvers Using SHRP2 Naturalistic Driving Data
Transportation Research Record | 2022
|Typical Acceleration Profiles for Left-Turn Maneuvers Based on SHRP2 Naturalistic Driving Data
SAE Technical Papers | 2021
|Driving Style Clustering using Naturalistic Driving Data
Transportation Research Record | 2019
|