Signal timing optimization at urban junctions is a multi-objective challenge critical to reducing congestion, travel time, and emissions in rapidly urbanizing cities. Conventional fixed-time and actuated signal plans in Nigeria, such as those deployed in Abuja, often fail to adapt to heterogeneous, non-lane-disciplined traffic flows. This study proposes a Differential Evolution (DE)-based optimization framework for dynamic signal control at the high- volume Wuse–Berger Junction, Abuja. Traffic volume, turning movement counts, queue lengths, and phase timing data were obtained through repeated manual classified counts conducted over multiple weekdays and weekends, capturing both peak and off-peak variations. The optimization model integrates adaptive phase sequencing with dynamic green time allocation, formulated as a multi-objective problem to minimize average delay, queue length, and estimated vehicular emissions. A hybrid parameter adaptation strategy is embedded within the DE algorithm to enhance convergence speed and avoid local optima. Performance was benchmarked against Particle Swarm Optimization (PSO), Genetic Algorithm (GA), and an actuated signal baseline within a calibrated VISSIM simulation model of the study site. Results show the DE-based controller achieved a 26.8% reduction in average delay and an 18.9% increase in throughput over the best-performing alternative. These findings underscore DE’s computational efficiency, scalability, and adaptability for intelligent traffic signal optimization, offering a viable pathway toward data-driven, context- sensitive control strategies in mixed-traffic urban environments.