ADVANCING TEMPORAL SAFETY PERFORMANCE FUNCTIONS: A COMPREHENSIVE EVALUATION OF EXPRESS LANES, RAMPS, AND RAMP METERING EFFECTS ON FREEWAY SAFETY

Abstract

Freeway safety remains a critical concern, especially in high-risk areas such as ramps, merges, and managed lane segments, where complex traffic interactions significantly elevate crash risks. This dissertation advances crash frequency prediction by developing short-term Safety Performance Functions (SPFs) that address the limitations of traditional long-term SPFs models and real-time safety analysis. By leveraging high-resolution microscopic traffic detector data from multiple states, the dissertation introduces innovative methodologies and delivers actionable insights into freeway safety dynamics. The dissertation pioneers the application of Multivariate Poisson-Lognormal (MVPLN) models to identify interdependencies between crashes at ramp and merge segments. To address challenges like data skewness and excessive zeros, advanced Bayesian frameworks, including the Negative Binomial Lindley (NB-L) and Poisson-Lognormal Lindley (PLN-L) models, are proposed. Additionally, a novel copula-based framework uncovers intricate safety relationships between Express Lanes (ELs) and General-Purpose Lanes (GPLs), offering new perspectives on inter-lane safety, particularly at critical access points. The key findings of the dissertation emphasize the significant role of traffic exposure, geometric configuration, and operational strategies in shaping crash risks. For instance, managed lanes, such as the I-4 Ultimate Express Lanes, exhibit unique safety patterns, with elevated crash risk associated with higher rightmost lane occupancy near merge areas. Ramp metering (RM) is demonstrated to effectively reduce crashes, particularly in weaving areas, with safety impacts varying based on control strategies and segment types. This dissertation delivers a robust framework for crash prediction and safety assessment, blending methodological advances with practical insights. Its contributions lay the groundwork for safer freeway designs, optimized Active Traffic Management (ATM) strategies, and enhanced safety practices. By bridging the gap between theoretical research and real-world applications, this dissertation equips policymakers, engineers, and researchers with the tools needed to improve freeway safety.