Addressing the core bottleneck of high-density operations in future Urban Air Mobility (UAM), this study investigates the optimization design method for airspace structure of UAV approach and departure procedures at vertiports. A structured multi-layer funnel-type hybrid airspace model is constructed, an airspace efficiency optimization function is designed, and genetic algorithms are employed for optimal airspace parameter design. Airspace optimization design calculations are performed in urban obstacle environments, and through comparative analysis with six typical airspace configurations, the designed airspace demonstrates significant balance in comprehensive performance. Results indicate that the strategy of simply increasing airspace dimensions to enhance capacity presents obvious performance trade-offs, while the compact columnar structure achieves coordinated multi-objective optimization by eliminating radial gradients and simplifying flight paths. The proposed genetic algorithm-based optimization design method for UAV approach and departure procedures in urban obstacle environments can effectively identify optimal parameter combinations and improve vertiport airspace operational efficiency.