以降低航班延误为目标的仿真优化方法研究

Simulation Optimization Approach to Reducing Flight Delays as the Objective

由于珠三角地区机场密集、空域紧张,导致该地区交通拥挤、管制员工作负荷大、运行安全隐患多、航班延误率高等。为此,迫切需要研究在不增加空域机场设施投资和保证运行安全的情况下降低该地区航班延误率的方法。考虑到难以对复杂的空中交通系统进行数学建模,采用空域机场仿真模型SIMMOD对珠三角地区空域机场系统进行仿真建模,并与并行扰动随机近似(SPSA,Simultaneous Perturbation Stochastic Approximation)优化算法相结合,以航班总延误成本最小为目标同时将管制员工作负荷限制在安全水平,优化航班时刻的仿真优化方法,降低该地区的航班延误率。采用THEIL不等式系数法对SIMMOD模型进行验证;采用非光滑反馈对SPSA算法进行改进,提高了算法的收敛速度和精度。仿真结果表明,采用优化的新航班时刻,可降低航班总延误成本38%;停机位延误、滑行道延误以及空中延误分别降低延误时间30%、26%和55%。

The Pearl River Delta Region （PRD） suffers the problem of high-density airports,airspace con- gestion, high controller workloads,large safety risks and a high delay of flights. It is necessary to develop a meth- od to decrease the flight delays with not increasing the investment cost of the airspace and airport infrastructure in PRD. Due to the complexity of mathematic modeling of the air traffic system,the SIMMOD simulation model of the PRD region＇s airspace and airport system is adopted in this paper,in combination with SPSA （Simultaneous Perturbation Stochastic Approximation） optimization algorithm,to minimize the cost of the total flight delays and keep the controller workload within the safety range. The simulation optimization of flight schedule is proposed to reduce the flight delays of this region. The SIMMOD model is validated by Theil＇s inequality coefficients. Using non-smooth feedback, SPSA optimization algorithm is improved to increase its rate of convergence and precision of the optimal solution. The simulation indicates that the optimum flight schedule can decrease effectively the cost of the total delays by 38% ,the gate delays by 30% ,the taxiway delays by 26% and the aerial delays by 55% re- spectively. The simulation optimization method has significant importance in reducing the overall delay cost and safety risks of PRD region＇s airspace and airport system.