Abstract

Collaborative Decision Making (CDM) embodies a new philosophy for managing air traffic. The initial implementation of CDM in the US has been aimed at Ground Delay Program Enhancements (GDP-E). However, the underlying concepts of CDM have the potential for much broader applicability. This paper reviews on-going and proposed CDM research streams. The topic areas discussed include: ground delay program enhancements; collaborative routing; performance monitoring and analysis; collaborative resource allocation mechanisms; game theory models for analyzing CDM procedures and information exchange; collaborative information collection and distribution. Document type: Part of book or chapter of book

Abstract

In this project the Ground Delay Program (GDP) and the Collaborative Decision Making (CDM) concepts have been studied. A statistical analysis on all GDPs that occurred during April of 2018 and during the year 2006 in the United States of America has been done. Some statistics have been derived on the number of GDPs and their causes. Results show that the 90% of the GDPs took place in only 13 and 16 airports, respectively. In both statistics Newark (EWR), in New York, have been the airport with more regulations. In addition, in April the 87% of the GDPs were caused by bad weather conditions versus a 74% in 2006. Furthermore, a complex study has been done at 2006 as the assigned delay and the number of affected flights per airport showing that the 90% of the total delay was assigned only in 10 airports, being Chicago the airport with more delay assigned. The same behaviour was observed with the number of affected flights: 90\\% of the flights has been related to only 10 different airports and Chicago again, was the airport with more affected flights. The results of the number of affected flights and the assigned delay have indicated a relationship. However, many airports with more flights and delay are also in the highest positions of the number of GDPs per airport list. The last part of the project is a simulation of as deterministic GHP for two different scenarios using Matlab. An optimal allocation of slots to flights taking into account its cost and the number of connection passengers of each flight has been analysed. The results have shown an exponential behaviour which increases for a high airport acceptance rates and when the flights have connection passengers. Using Ratio-by-Schedule the minimum costs have been obtained.

Abstract

We develop and evaluate two significant modeling concepts within the context of a large-scale Airspace Planning and Collaborative Decision-Making Model (APCDM) and, thereby, enhance its current functionality in support of both strategic and tactical level flight assessments. The first major concept is a new severe weather-modeling paradigm that can be used to assess existing tactical en route flight plan strategies such as the Flight Management System (FMS) as well as to provide rerouting strategies. The second major concept concerns modeling the mediated bartering of slot exchanges involving airline trade offers for arrival/departure slots at an arrival airport that is affected by the Ground Delay Program (GDP), while simultaneously considering issues related to sector workloads, airspace conflicts, as well as overall equity concerns among the airlines. This research effort is part of an $11.5B, 10-year, Federal Aviation Administration (FAA)-sponsored program to increase the U.S. National Airspace (NAS) capacity by 30 percent by the year 2010. Our innovative contributions of this research with respect to the severe weather rerouting include (a) the concept of â Probability-Netsâ and the development of discretized representations of various weather phenomena that affect aviation operations; (b) the integration of readily accessible severe weather probabilities from existing weather forecast data provided by the National Weather Service (NWS); (c) the generation of flight plans that circumvent severe weather phenomena with specified probability levels, and (d) a probabilistic delay assessment methodology for evaluating planned flight routes that might encounter potentially disruptive weather along its trajectory. Given a fixed set of reporting stations from the CONUS Model Output Statistics (MOS), we begin by constructing weather-specific probability-nets that are dynamic with respect to time and space. Essential to the construction of the probability-nets are the point-by-point forecast probabilities associated with MOS reporting sites throughout the United States. Connections between the MOS reporting sites form the strands within the probability-nets, and are constructed based upon a user-defined adjacency threshold, which is defined as the maximum allowable great circle distance between any such pair of sites. When a flight plan traverses through a probability-net, we extract probability data corresponding to the points where the flight plan and the probability-net strand(s) intersect. The ability to quickly extract this trajectory-related probability data is critical to our weather-based rerouting concepts and the derived expected delay and related cost computations in support of the decision-making process. Next, we consider the superimposition of a flight-trajectory-grid network upon the probability-nets. Using the U.S. Navigational Aids (Navaids) as the network nodes, we develop an approach to generate flight plans that can circumvent severe weather phenomena with specified probability levels based on determining restricted, time-dependent shortest paths between the origin and destination airports. By generating alternative flight plans pertaining to specified threshold strand probabilities, we prescribe a methodology for computing appropriate expected weather delays and related disruption factors for inclusion within the APCDM model. We conclude our severe weather-modeling research by conducting an economic benefit analysis using a k-means clustering mechanism in concert with our delay assessment methodology in order to evaluate delay costs and system disruptions associated with variations in probability-net refinement-based information. As a flight passes through the probability-net(s), we can generate a probability-footprint that acts as a record of the strand intersections and the associated probabilities from origin to destination. A flight planâ s probability-footprint will differ for each level of data refinement, from whence we construct route-dependent scenarios and, subsequently, compute expected weather delay costs for each scenario for comparative purposes. Our second major contribution is the development of a novel slot-exchange modeling concept within the APCDM model that incorporates various practical issues pertaining to the Ground Delay Program (GDP), a principal feature in the FAAâ s adoption of the Collaborative Decision-Making (CDM) paradigm. The key ideas introduced here include innovative model formulations and several new equity concepts that examine the impact of â at-least, at-mostâ trade offers on the entire mix of resulting flight plans from respective origins to destinations, while focusing on achieving defined measures of â fairnessâ with respect to the selected slot exchanges. The idea is to permit airlines to barter assigned slots at airports affected by the Ground Delay Program to their mutual advantage, with the FAA acting as a mediator, while being cognizant of the overall effect of the resulting mix of flight plans on air traffic control sector workloads, collision risk and safety, and equity considerations. We start by developing two separate slot-exchange approaches. The first consists of an external approach in which we formulate a model for generating a set of package-deals, where each package-deal represents a potential slot-exchange solution. These package-deals are then embedded within the APCDM model. We further tighten the model representation using maximal clique cover-based cuts that relate to the joint compatibility among the individual package-deals. The second approach significantly improves the overall model efficiency by automatically generating package-deals as required within the APCDM model itself. The model output prescribes a set of equitable flight plans based on admissible trades and exchanges of assigned slots, which are in addition conformant with sector workload capabilities and conflict risk restrictions. The net reduction in passenger-minutes of delay for each airline is the primary metric used to assess and compare model solutions. Appropriate constraints are included in the model to ensure that the generated slot exchanges induce nonnegative values of this realized net reduction for each airline. In keeping with the spirit of the FAAâ s CDM initiative, we next propose four alternative equity methods that are predicated on different specified performance ratios and related efficiency functions. These four methods respectively address equity with respect to slot-exchange-related measures such as total average delay, net delay savings, proportion of acceptable moves, and suitable value function realizations. For our computational experiments, we constructed several scenarios using real data obtained from the FAA based on the Enhanced Traffic Management System (ETMS) flight information pertaining to the Miami and Jacksonville Air Route Traffic Control Centers (ARTCC). Through our experimentation, we provide insights into the effect of the different proposed modeling concepts and study the sensitivity with respect to certain key parameters. In particular, we compare the alternative proposed equity formulations by evaluating their corresponding slot-exchange solutions with respect to the net reduction in passenger-minutes of delay for each airline. Additionally, we evaluate and compare the computational-effort performance, under both time limits and optimality thresholds, for each equity method in order to assess the efficiency of the model. The four slot-exchange-based equity formulations, in conjunction with the internal slot-exchange mechanisms, demonstrate significant net savings in computational effort ranging from 25% to 86% over the original APCDM model equity formulation. The model has been implemented using Microsoft Visual C++ and evaluated using a C++ interface with CPLEX 9.0. The overall results indicate that the proposed modeling concepts offer viable tools that can be used by the FAA in a timely fashion for both tactical purposes, as well as for exploring various strategic issues such as air traffic control policy evaluations; dynamic airspace resectorization strategies as a function of severe weather probabilities; and flight plan generation in response to various disruption scenarios. Ph. D.

Abstract

This dissertation investigates predictability in strategic air traffic management with a focus on ground delay programs (GDPs). Through a survey of flight operators, we confirm the proposition that flight operators care about predictability. We then develop models that incorporate predictability into GDP cost optimization after failing in finding an existing model that can serve this purpose. This is accomplished by modifying traditional GDP delay cost functions so that they incorporate predictability, and determining the sensitivities of the optimal planned capacity recovery time and associated cost to the unpredictability premiums included in the cost functions. To do this, we develop two stochastic GDP models: a GDP no-revision, or static, model; and a GDP revision, or dynamic, model considering one GDP revision. GDP scope, which matters only in the revision model, is also considered. The optimization results from the case study show that the cost of unpredictability clearly matters, particularly in the more realistic case where GDP revision is allowed. Of the two unpredictability cost parameters, the one for unplanned delay has a stronger impact than the one for planned un-incurred delay. The insights from this analysis might eventually be used to develop a decision support tool that air traffic managers could use in determining what the planned end time should be for a GDP in a manner that reflects the importance of predictability to flight operators.