Metaheuristics are high-level problem-independent algorithmic frameworks for solving difficult optimization problems. They provide strategies to explore search spaces efficiently, balancing intensification (exploiting good solutions) and diversification (exploring new regions). Common metaheuristics include: Genetic Algorithms, Simulated Annealing, Tabu Search, Ant Colony Optimization, Particle Swarm Optimization, and Variable Neighborhood Search. Characteristics: (1) do not guarantee optimal solutions but find good solutions in reasonable time; (2) applicable to black-box problems; (3) suitable for large-scale and NP-hard problems; (4) often outperform exact methods on real-world instances. Design involves: solution representation, neighborhood structure, evaluation function, and perturbation mechanisms.