关键词: 空中交通, 多机场系统, 仿真优化, 并行扰动随机近似, 非光滑反馈

Abstract: Pearl River Delta region（PRD） suffers the problem of high-density airports, airspace congestion and a high delay-rate of flights. Based on a delay hierarchical model as the optimized target, combined the SIMMOD simulation model of the PRD region’s multi-airport system and restricted SPSA（Simultaneous Perturbation Stochastic Approximation） optimization algorithm, the simulation optimization of flight schedule in the PRD region is discussed to reduce the flight delay of this region. SPSA optimization algorithm is improved using non_smooth feedback. The simulation indicates that the optimum flight schedule can decrease the total delay effectively. The method proposed has significant importance on reducing the overall delay cost of PRD region’s multi-airport system.

Key words: air traffic, multi-airport system, simulation optimization, simultaneous perturbation stochastic approximation, non_smooth feedback