Modeling, Optimization, and Operation of Large-Scale Air Traffic Flow Management on Spark

Chen, Jun; Cao, Yi; Sun, Dong

doi:10.2514/1.i010533

Cited by 11 publications

(3 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The opening scheme of operating sectors L τ must be taken into account in which capacity is a function of time (refer to [16] for more details of how the dynamic capacity is modelled). Finally, Constraints (11) and (12) specify the binary constraints. For more details about this formulation, the reader may refer to [16].…”

Section: Original Dcb Optimisermentioning

confidence: 99%

“…Then, a big data processing model named Hadoop MapReduce was used in [11] to spread computationally intensive tasks to Hadoop clusters for concurrent executions. Further, a novel layered aggregate model was developed in [12] for handling flexible rerouting problem, which was not well handled in the previous link transmission model. The model can be efficiently computed by a fast cluster architecture as provided by Spark, reported twice as fast as the Hadoop MapReduce model.…”

mentioning

confidence: 99%

See 1 more Smart Citation

Fast-Time Demand-Capacity Balancing Optimizer for Collaborative Air Traffic Flow Management

Camargo

Prats

2021

Journal of Aerospace Information Systems

View full text Add to dashboard Cite

I. IntroductionUnder current air traffic management (ATM) operations, one of the primary causes for flight delays and congestion is that the number of flights (demand) often exceeds the supply of the airspace accommodation (capacity). The effort thereby to achieve demand and capacity balancing (DCB) is typically known as air traffic flow management (ATFM), which is regarded as an enabler of ATM efficiency and effectiveness [1]. Since the last few decades, a number of researchers have focused their activity on the development of network ATFM models to minimise the congestion costs incurred from demand and capacity imbalances in both airports and airspace volumes (sectors). However, it is important to highlight that the existing network ATFM models present significant challenges in computational tractability when dealing with large-scale problems. As pointed out in [2], this could be one of the main reasons, besides the fairness concern, why they have not been successfully transitioned into practice.In response to this issue, the computational challenges have been targeted by researchers through applying different numerical computation methods. For the models based on the well-studied Bertsimas Stock-Patterson model [3], a Dantzig-Wolfe decomposition method was proved highly efficient to solve the block-angular formulation [4,5], where the network traffic flow can be decomposed flight by flight. As reported in [6], the effects of this parallel speed-up were further improved by an implementation of the Graphics Processing Units (GPUs), due to the large amount of cores over the Central Processing Units (CPUs). A similar column generation approach was also adopted in [7], where ground holds, airborne delays, reroutes and cancellations are considered as possible control actions, and the approach was then extended to generate recourse strategies in the presence of uncertain capacity constraints.With regards to the network ATFM models that are based on Eulerian formulation, a dual-decomposition method was used, which decomposes the traffic flow path by path, and each flight path was solved (optimised) independently [8]. The dimension of a large-scale cell transmission model therefore depends only on the number of identified flight

show abstract

Section: Original Dcb Optimisermentioning

confidence: 99%

mentioning

confidence: 99%

Fast-Time Demand-Capacity Balancing Optimizer for Collaborative Air Traffic Flow Management

Camargo

Prats

2021

Journal of Aerospace Information Systems

View full text Add to dashboard Cite

show abstract

“…Chen, Hu, Zhang, Yin, and Han [29], Chen, Cao, and Sun [31], and Wei, Cao, and Sun [113]. The difference between this approach and the previous works is that the usage of commodities does not specify to which particular flights are assigned the proposed modifications in the flight plans.…”

Section: Literature Reviewmentioning

confidence: 99%