<p>Fog computing supports low-end-to-end latency and location-aware services for Internet of Things (IoT) applications by extending cloud capabilities to the network edge. However, task scheduling in fog environments faces significant challenges due to resource heterogeneity, energy constraints, and conflicting quality-of-service requirements such as execution delay and energy consumption. Existing scheduling algorithms often focus on single objectives, failing to achieve an effective trade-off in end-to-end latency-sensitive and energy-critical scenarios like intelligent transportation. To address this issue, this paper proposes a Fuzzy Chaos Genetic Algorithm (FCGA) to optimize multi-objective task scheduling in fog computing. FCGA enhances the traditional genetic algorithm by integrating chaos theory for population initialization and perturbation, thereby improving solution diversity and avoiding premature convergence. Additionally, a fuzzy logic controller is designed to dynamically adjust crossover and mutation rates, balancing global exploration and local exploitation. Joint energy and execution delay awareness are integrated into the fitness evaluation to support balanced optimization of conflicting objectives. Extensive simulations are conducted in an intelligent traffic management scenario under varying workloads and node scales. Experimental results show that FCGA achieves superior performance in terms of makespan, execution delay, energy consumption, and network utilization compared to baseline approaches, demonstrating its ability to balance critical performance metrics including response time and power consumption under dynamic fog environments.</p>