基于大规模并行D-W分解算法的空中交通短期流量管理模型研究

Massive Parallel Dantzig-Wolfe Decomposition Applied to Traffic Flow Scheduling

针对空中交通短期流量管理问题,给出了整机型的空中交通短期流量管理模型(ST-TFMP)。运用大规模并行的D-W分解算法,结合数据仓储技术,快速解决了航班延误最小问题。同时对算法的计算性能进行了计算估计和实验分析。通过对实际终端区流量进行仿真计算,仿真结果验证了本文提出的并行算法的合理性。

For the problem of short-term air traffic flow management we give the short-term air traffic flow management model （ST-TFMP） for individual flights. Then by using massively parallel D- W decomposition algorithm which is combined with data warehousing technology, the problem of minimum flight delays is solved in a short time. At the same time, the computation performance is evaluated by experiment. Through the simulation calculation of accrual terminal area traffic flow, the simulation result validates the rationality of the parallel algorithm.