Driver Performance and Behavior in Adverse Weather Conditions: Microsimulation and Variable Speed Limit Implementation of the SHRP2 Naturalistic Driving Study Results - Phase 3
-
2021-10-01
-
Details:
-
Creators:
-
Corporate Creators:
-
Corporate Contributors:Wyoming. Department of Transportation ; United States. Department of Transportation. Federal Highway Administration ; American Association of State Highway and Transportation Officials ; United States. Transportation Research Board ; Transportation Research Board. Strategic Highway Research Program 2
-
Subject/TRT Terms:
-
Publication/ Report Number:
-
Resource Type:
-
Geographical Coverage:
-
Edition:October 2021, Phase 3 Final Report
-
Contracting Officer:
-
Corporate Publisher:
-
Abstract:The negative impact of adverse weather conditions on traffic safety and operation is the primary focus of many studies. However, in-depth investigations of driver behavior and performance during adverse weather at a trajectory level using naturalistic driving data are limited. Over the years, researchers have utilized aggregate traffic and weather data to investigate their impact on roadway network operations and safety; however, these data might not represent the actual driving environment at a trajectory level. The novel approaches presented in this report, could fill existing gaps by investigating the safety and operational impacts of human behavior in conjunction with other factors related to weather, traffic, and roadway geometry via disaggregate trajectory-level data available through the Second Strategic Highway Research Program (SHRP2), Naturalistic Driving Study (NDS) and Roadway Information Databases. First, parametric and non-parametric models were utilized to better understand different behavioral factors, including lane-keeping, lane changes, gap acceptance, and speed selection, in adverse weather conditions. Findings from the behavioral investigation could be integrated into Weather-responsive Active Traffic Management (ATM) to disseminate safety messages in adverse weather conditions. Afterward, a unique radar-vision algorithm was developed to process trajectory-level data from instrumented vehicles to continuously predict driving states and estimate driving events using trips from the SHRP2 NDS. Subsequently, emphasis has been provided on developing reliable, accurate, and efficient machine learning-based lane change detection and prediction models through a data fusion approach considering different data availability. Additionally, essential indicators of near-crashes and the zone of interest for surrogate measures of safety were identified and thoroughly investigated. The lack of studies investigating driver behavior in adverse weather conditions stem from the fact that weather identification using video data is a daunting task. Therefore, several cost-effective and reliable detection systems were developed to detect trajectory-level weather information at a road-surface level in real-time using cutting-edge machine and deep learning techniques. Through crowdsourcing, real-time weather information can be integrated with Traffic Management Centers (TMCs) via Connected Vehicle Technologies. Finally, findings from the SHRP2 NDS were leveraged to develop Weather-based Microsimulation models to assess the safety and operational performance of ITS-based countermeasures on Wyoming freeways including Variable Speed Limit systems and the Wyoming Connected Vehicle Pilot Program. This study unlocked new horizons and potentials in conducting adverse weather-related research utilizing the NDS data, providing unprecedented opportunities to improve the reliability and effectiveness of safety and operational countermeasures.
-
Format:
-
Funding:
-
Collection(s):
-
Main Document Checksum:
-
Download URL:
-
File Type: