Akademik Veri Yönetim Sistemi
IATSS Research, cilt.50, sa.2, ss.882-897, 2026 (ESCI, Scopus)
Priority intersections are among the most collision-prone elements of urban road networks, yet geometric design decisions are often evaluated for safety and operational capacity separately. In the UK, design requirements have shifted from larger, motor-traffic-oriented geometric standards toward lower values, particularly in residential contexts, reflecting safety evidence linking geometry to collision occurrence. However, while this has improved safety-focused design guidance, reduced geometric values may constrain capacity, and existing studies do not provide a framework for identifying geometric designs that jointly balance safety and capacity. To address this gap, this study develops an optimization framework for jointly evaluating priority intersection geometry in terms of safety and capacity. The framework integrates collision prediction models with capacity formulations and applies a genetic algorithm to identify feasible geometric configurations that improve both outcomes. The methodology was applied to 120 priority intersections in Portsmouth using real-world geometric, traffic, and collision data. The optimization generated alternative layouts for each site and, in aggregate, reduced the predicted number of collisions by 43% compared with existing designs, while improving operational efficiency under relevant traffic conditions. Corner radii and major road approach widths were the most influential variables, with effects consistent with, but also refining, previous literature by showing that safety benefits depend on speed conditions and traffic demand. The results indicate that modest geometric adjustments can substantially reduce collisions at priority intersections without compromising operational performance, offering a practical basis for enhancing intersection design practice and decision-making frameworks.