No description available.

Cette thèse a pour but de modéliser et d'optimiser deux composantes du système de gestion de flux de trafic aérien : la prédictibilité et la flexibilité. Cette modélisation est équivalente à établir une relation entre la fenêtre temporelle et les taux d'arrivée des avions. Deux approches sont utilisées : l'analyse des données historiques et la modélisation mathématique. L'analyse des données historique a permis de déterminer la fenêtre temporelle raisonnable mais sans pouvoir apporter les améliorations nécessaires pour y arriver. La modélisation mathématique permet non seulement de définir de façon rigoureuse la prédictibilité et la flexibilité mais également de traiter des vols en différents scénarios de priorités. La combinaison de DC algorithme avec des méthodes de résolutions classiques comme Branch and Bound a nettement amélioré la vitesse de la convergence des solutions et donc elle peut être utilisée pour la phase tactique de gestion de flux du trafic aérien. / This thesis aims to model and optimise two components of the air traffic flow management system : predictibility and flexibility. This modelling is equivalent to establishing a relationship between the time window and the rate of arrival flights. Two approachs are used : the analysis of historical data and mathematical modeling. The analysis of historical data was used to establish the relationship between the time window and arrivla rate of flights. It provided the optimal time window but could not show how to modify the system to lead to that time window. Mathematical modeling can not only define the predictability and flexibility in the rigourous manner but also deal with different scenarios of fligths priorities. The combination of DC algorithm with classical methods like Branch and Bound has significantly improved the speed of convergence of solutions and therefore it can be used for the tactical phase of the air traffic flow management.

The Federal Aviation Administration (FAA) has been modernizing the United States' air transportation system within a series of initiatives called the Next Generation Air Transportation System (NextGen). The goal of NextGen is to increase the safety, efficiency, capacity, predictability, and resiliency of American Air Traffic Control (ATC) by implementing satellite-based communication, and navigation systems. Because of the vast oceanic areas controlled by Oakland, New York, and Anchorage air traffic control centers, improving oceanic operations is significant for the United States. According to the FAA, oceanic flights generate 31% of passenger revenue and 40% of cargo revenue in U.S.-controlled airspace. New NextGen procedures offer the opportunity for aircraft to save fuel consumption by allowing oceanic flights to fly at more efficient routes and flight levels. This dissertation investigates three areas to improve flight operations over oceanic airspace. The first area studies the operational benefits of providing satellite-based meteorological information to aircraft operating in remote and oceanic airspace. This research effort uses two approaches as follows: 1) statistical flight analysis, and 2) simulation-based analysis. The second area provides an optimization technique to improve the current procedures for assigning flights to the Organized Track System (OTS) in the Atlantic Ocean based on the Collaborative Decision Making (CDM) concept. The third area investigates the potential savings of "In-Trail Procedure" (ITP) as one of the advanced surveillance operations in the Pacific and Atlantic oceanic airspace. To quantify the operational benefits of the proposed procedures, a fast-time simulation tool, the Global Oceanic (GO) model, is developed and employed. The GO model is a microscopic flight simulation tool that has been developing by the Air Transportation Systems Laboratory at Virginia Tech offering realistic and inexpensive evaluations of novel technologies and procedures to improve flight operations over global oceanic airspace. the results of these studies are analyzed in terms of fuel consumption, travel distance, travel time, level of service, and potential air traffic controllers' workload. / Doctor of Philosophy / The economic growth and social connectivity of nations are highly correlated to effective and efficient air transportation systems. The Federal Aviation Administration (FAA) has initiated a program to modernize America's air transportation system and make flight operations safer, and more efficient. This program is called the Next Generation Air Transportation System (NextGen) and its goal is transforming the communication and navigation technologies to satellite-based systems. Improving oceanic flights is one of the main concerns of the NextGen program since the United States controls massive oceanic areas in the Atlantic and the Pacific Ocean. The FAA needs to evaluate the benefits and costs of advanced technologies and procedures to justify the NextGen initiatives. The FAA has employed computer simulation tools to support decisions for future infrastructure investments and encourage airlines to equip their aircraft with more advanced avionics. The Global Oceanic (GO) model is a microscopic flight simulation tool developed jointly by the Air Transportation Systems Laboratory at Virginia Tech and the FAA providing quick, realistic, and inexpensive evaluations of advanced procedures to improve flight operations over oceans. This dissertation investigates the operational benefit of three advanced procedures using the GO model. The areas to improve flight operations over oceanic airspace are as follows: 1) operational benefits of providing satellite-based meteorological information to aircraft operating in remote and oceanic airspace, 2) operational benefits of an optimization technique for flight assignments to the Organized Track System (OTS) in the Atlantic Ocean, 3) operational benefits of "In-Trail Procedure" (ITP) as one of the advanced surveillance operations in the Pacific and Atlantic oceanic airspace. These studies quantify the potential savings of these procedures in terms of reducing fuel consumption, travel distance, travel time, greenhouse gas emissions, and potential air traffic controllers' workload.

Der Luftverkehr stellt ein komplexes Gesamtsystem dar, in welchem eine Prozessoptimierung aufgrund zahlreicher und verschiedenartiger Arbeitsabläufe verschiedener Unternehmen nur durch eine übergeordnete Leistungsbewertung möglich ist. Hierfür wurde im Bereich des leistungsbasierten Flughafenmanagements - sowohl auf wissenschaftlicher, als auch industrieller Ebene - bereits eine hohe Zahl individueller Lösungsansätze entwickelt, die jedoch aufgrund ihrer unterschiedlichen Struktur schwer zu vergleichende Ergebnisse hervorbringen. Des Weiteren werden Wechselbeziehungen zwischen ausschließlich eindimensional betrachteten Leistungsindikatoren, welche die verschiedenen Prozessschritte einzeln abbilden, außer Acht gelassen - die Dynamik des Systems spiegelt sich auf diese Art nicht auf Datenebene wider. Abhängigkeiten beeinflussen jedoch maßgeblich das Bewertungsergebnis und sind damit bei der Implementierung von Optimierungsstrategien innerhalb heutiger Konzepte essentiell. Der Kern dieser Dissertation umfasst die erweiterte datenbasierte Betrachtung des Luftverkehrssystems innerhalb des leistungsbasierten Ansatzes und einer damit einhergehenden Entkopplung von Modellierungsansätzen. Dies bedeutet, dass das betrachtete Luftverkehrssystem nur durch Leistungskennwerte beschrieben werden soll (z. B. Verspätungen, meteorologische Messungen). Der Einsatz künstlicher neuronaler Netze erhöht die Möglichkeiten zur Abbildung und Erfassung nicht-linearer und abhängiger Wert-diskreter Zeitreihen, welche im ständigen Vergleichsprozess wesentlich für die Strategiebildung sind. Die Datenquellen beziehen sich einerseits auf mikroskopische Analysen im Bereich des Boardings und damit verbundenen wissenschaftlichen Ausarbeitungen, andererseits auf die Beispielflughäfen Flughafen Hamburg (HAM) und Gatwick Airport (LGW) der Jahre 2012 - 2015 im makroskopischen Fokus. Die Implementierung des Wetters erfolgt über aggregierte meteorologische Kennzahlen, welche auf realen Wettermeldungen des entsprechenden Zeitraums beruhen. Von der Wahl und Definition eines Systems (Boarding, HAM und LGW) ausgehend, erfolgt eine geeignete Datenaggregation, welche Daten zur anschließenden Wissensgenerierung bereitstellt und damit Optimierungsansätze ermöglicht. Im Sinne des wachsenden Interesses der Forschung im Bereich des leistungsbasierten Luftverkehrsmanagements und der heutigen Wichtigkeit von Entscheidungsunterstützungssystemen bei der Strategieentwicklung, fokussiert sich diese Arbeit damit auf die Durchführung multivariater nicht-linearer Zeitreihenanalyse und -vorhersage mit neuronalen Netzen. Die damit einhergehende Nachvollziehbarkeit solcher Datenreihen liefert Möglichkeiten zur Systemidentifikation (datenbasiertes Erlernen der Systemdynamik). Das identifizierte Systemabbild des Luftverkehrs kann folglich für Simulation bekannter Eingabegrößen, als auch für die optimierte Kontrolle des Systems herangezogen werden und umfasst damit wesentliches Erweiterungspotential für heutige Management-Konzepte, um Entwicklungen hin zu einem kooperativen Betrieb zu unterstützen. Ableitend aus der Differenzierung in mehrere gekoppelte Bearbeitungsschritte innerhalb dieser Arbeit, erfolgt eine Fokussierung auf drei Kernfragen: a) Ist eine auf Leistungskennwerten des Luftverkehrs basierende Systemidentifikation mit derzeitigen Paradigmen neuronaler Netze möglich? b) Welche Einschränkungen sind gemäß des unterschiedlichen Charakters der Datensätze zu beachten und wie kann diesen durch eine, das Training der neuronalen Netze vorbereitenden, Datenstrukturanalyse und -Anpassung entgegengewirkt werden? c) Ist auf Basis der trainierten Netze eine erweiterte Optimierung und Vorhersage innerhalb vorhandener Strukturen des leistungsbasierten Luftverkehrsmanagements möglich?:I Grundlagen 1 Leistungsbasiertes Luftverkehrsmanagement 1.1 Allgemeine Definitionen & Begrifflichkeiten 1.2 Leistungsbewertungsrahmenwerke des Luftverkehrs 1.2.1 Interdependenzen von Leistungsindikatoren 1.2.2 Einflussfaktoren auf die Leistungsfähigkeit von Verkehrsflughäfen 1.2.3 Flugmeteorologische Datenaggregation 1.3 Grundkonzepte leistungsbasierten Luftverkehrsmanagements 1.3.1 Airport Collaborative Decision Making (A-CDM) 1.3.2 Total Airport Management (TAM) 1.3.3 Performance Based Airport Management (PBAM) 2 Künstliche neuronale Netze 2.1 Grundverfahren der Computational Intelligence 2.2 Biologische neuronale Netze 2.3 Topologie & Bestandteile künstlicher neuronaler Netze 2.4 Lernvorgang & Fehlerevaluierung 2.5 Netzparadigmen 2.5.1 Feedforward Netzwerke 2.5.2 Rekurrente (rückgekoppelte) Netzwerke 2.5.3 Faltende Netzwerke 2.6 Taxonomie der Zeitreihenverarbeitung mit künstlichen neuronalen Netzen 2.6.1 Zeitreihenregression 2.6.2 Zeitreihenklassifikation II Anwendung künstlicher neuronaler Netze im Luftverkehr 3 Methodische Konzeption 3.1 Datenbasierte Erfassung des Luftverkehrssystems 3.1.1 Prinzip des virtuellen Luftverkehrssystems 3.1.2 Systemidentifikation, Kontrolle & Simulation 3.2 Anwendungsgebiet Turnaround Management 3.2.1 Status Quo 3.2.2 Stochastische Boardingsimulation & Aggregationsmetrik 3.3 Anwendungsgebiet Air Traffic Flow Management 3.3.1 Status Quo 3.3.2 Datengrundlage der Flughäfen Hamburg & London Gatwick 4 Systemidentifikation mit künstlichen neuronalen Netzen 4.1 Modularisierung der Systemidentifikation 4.2 Problemidentifikation & Systemauswahl 4.3 Datenstrukturanalyse & -anpassung 4.3.1 Datenvoranalyse 4.3.2 Datenvorverarbeitung & Fehlerbereinigung 4.3.3 Datenanpassung für maschinelles Lernen 4.4 Paradigmenauswahl & Modellinitialisierung 4.5 Modellanwendung & -Überwachung 4.6 Nachgelagerte Analyse & Modellanpassung 5 Simulation & Kontrolle durch künstliche neuronale Netze 5.1 Sequenzprediktion zur Vorhersage bei bekannten Eingabemengen 5.1.1 Fehlerfortschreitung & Prognoseevaluierung 5.1.2 Robustheitsschätzung der Vorhersage 5.2 Nutzung des Modells als adaptive Kontrollstruktur 5.2.1 Extraktion von Zusammenhängen der Eingabegrößen 5.2.2 Metaheuristische Optimierung der Eingabemengen III Anwendung 6 Anwendungsgebiet Boarding als Teil des Turnaround-Prozesses 6.1 Systemidentifikation des Boarding-Subsystems 6.1.1 Datenstrukturanalyse 6.1.2 Experimenteller Aufbau & Modellanwendung 6.2 Simulation & Kontrolle des Boarding-Subsystems 6.2.1 Vorhersage des Boardingvorgangs unter verschiedenen Randbedingungen 6.2.2 Robustheitsprüfung der adaptiven Kontrollstruktur 6.2.3 Ableitung von Handlungsempfehlungen innerhalb des Prozesses 7 Anwendungsgebiet Air Traffic Flow Management 7.1 Systemidentifikation des Flugbetriebs in Hamburg & London Gatwick 7.1.1 Datenstrukturanalyse 7.1.2 Experimenteller Aufbau & Modellanwendung 7.2 Simulation & Kontrolle des Flugbetriebs in Hamburg & London Gatwick 7.2.1 Vorhersage von Verspätungen unter Berücksichtigung des Wetters 7.2.2 Robustheitsprüfung der adaptiven Kontrollstruktur 7.2.3 Eingabeempfehlungen zur Minimierung der Gesamtverspätung 8 Schlussbetrachtungen 8.1 Zusammenfassende Auswertung der Basisszenarien 8.2 Ausblick

As air traffic congestion grows, air traffic flow management (ATFM) is becoming a great concern. ATFM deals with air traffic and the efficient utilization of the airport and airspace. Air traffic efficiency is heavily influenced by unanticipated factors, or uncertainties, which can come from several sources such as mechanical breakdown; however, weather is the main unavoidable cause of uncertainty. Because weather is unpredictable, it poses a critical challenge for ATFM in current airport and airspace operations. Convective weather results in congestion at airports as well as in airspace sectors. During times of congestion, the decision as how and when to send aircraft toward an airspace sector in the presence of weather is difficult. To approach this problem, we first propose a two-stage stochastic integer program by emphasizing a given single sector. By considering ground delay, cancellation, and cruise speed for each flight on the ground in the first stage, as well as air holding and diversion recourse actions for each flight in the air in the second stage, our model determines how aircraft are sent toward a sector under the uncertainty of weather. However, due to the large number of weather scenarios, the model is intractable in practice. To overcome the intractability, we suggest a rolling horizon method to solve the problem to near optimal. Lagrangian relaxation and subgradient method are used to justify the rolling horizon method. Since the rolling horizon method can be solved in real time, we can apply it to actual aircraft schedules to reduce the costs incurred on the ground as well as in airspace. We then extend our two-stage model to a multistage stochastic program, which increases the number of possible weather realizations and results a more efficient schedule in terms of costs. The rolling horizon method as well as Lagrangian relaxation and subgradient method are applied to this multistage model. An overall comparison among the previously described methodologies are presented.

In ongoing efforts to balance air traffic demand and airspace capacity, airspace complexity stands as a fundamental research problem. Taking a more analytic approach, this thesis proposes that airspace complexity can be described in terms of how the airspace (together with the traffic inside it and the traffic control algorithm) responds to disturbances. The response of the airspace to a disturbance is captured by the degree of control activity required to accommodate such disturbance. Furthermore, since the response of the airspace depends on the disturbance, this thesis introduces a complexity map which shows how an airspace responses to a set of different disturbances. Among the many possible types of disturbances, this thesis considers an aircraft entering into the airspace, and the proposed method of describing airspace complexity is illustrated with examples. The time evolution of a complexity map is investigated using a statistical approach. In addition, the proposed method is illustrated in relation to current and future traffic flow management concepts. It is also shown that the proposed method can be applied to airspace design problems.

Today, the information (signal timings, detector extension, phase sequence, etc.) to install traffic lights on the street are obtained from traffic software simulations platforms, meaning that information from simulation is not tested on the field (intersection where it will be installed) before the installation. Many installed controllers on the street use time of day (TOD) patterns due to cheaper cost than adaptive traffic control systems, but that is not the best solution for traffic volume changes that can occur during the day or even a month. To improve traffic signal operation most of the traffic signal controllers in the same corridor or zone operate in coordination mode. Furthermore, phases need to be in coordination to achieve “green wave”. Green wave is term used when in corridor traffic lights allow continues flow of traffic through intersections that are coordinated. / Includes bibliography. / Thesis (M.S.)--Florida Atlantic University, 2016. / FAU Electronic Theses and Dissertations Collection

Most U.S. urban traffic signal systems deploy multiple signal timing plans to account for daily variability of traffic demand (i.e. morning peak, midday, afternoon peak, off peak and night). Groups of signals (belonging to the one zone or section) along an urban arterial, usually operate in a coordinated manner. This essentially means that timing plans change at the same time for all the signals in the group, so as to facilitate vehicle progression of through a series of signals. Good traffic signal timing practices assume a certain level of monitoring and maintenance in order to guarantee that they are efficient in servicing current traffic conditions. / Includes bibliography. / Thesis (M.S.)--Florida Atlantic University, 2015 / FAU Electronic Theses and Dissertations Collection

In recent years, there has been a considerable increase in the amount of construction work on the U.S. national highways. Most of the work undertaken is the reconstruction and rehabilitation of the existing transportation networks. Work zones in the United States are likely to increase in number, duration and length due to emphasis on repair and highway reconstruction as a significant portion of all federal-aid highway funds are now geared toward highway rehabilitation. The challenge of mobility is particularly acute in work zone areas as road repair and construction intensifies traffic issues and concentrates them in specific locations and at specific times. Due to the capacity drop, which is the result of lane closure in work zone area, congestion will occur with a high traffic demand. The congestion increases number and severity of traffic conflicts which raise the potential for accidents; furthermore traffic operational properties of roadway in work zone area become worse. Intelligent Transportation System (ITS) technologies have been developed and are being deployed to improve the safety and mobility of traffic in and around work zones. In several states in the US, the use of Dynamic Merge Controls also known as Dynamic Lane Merge (DLM) system has been initiated to enhance traffic safety and to improve traffic flow in work zone areas. The DLM usually takes two forms; dynamic iii early merge and dynamic late merge. The use of variable speed limit (VSL) systems at work zones is also one of those measures. VSL systems improve safety by helping the driver in determining the maximum speed that drivers should travel. Besides adding improvement to safety, they are also expected to improve mobility at the work zones. The main goal of this study is to evaluate the safety and operational effectiveness of the dynamic merge systems i.e. the dynamic early lane merge and dynamic late lane merge, in the presence of VSL system. More specifically, the VISSIM model is utilized to simulate a twoto-one lane configuration when one out of the two lanes in the work zone is closed for traffic. Six different scenarios were adopted to assess the effectiveness of these scenarios under different traffic demand volumes and different drivers‟ compliance rates to the messages displayed by the systems. These scenarios are;  Work Zone without VSL and without SDLMS or the current Motorist Awareness System (MAS)  Work Zone with VSL and without SDLMS  Work Zone with VSL and Early SDLMS  Work Zone with VSL and Late SDLMS  Work Zone with early SDLMS and without VSL  Work Zone with late SDLMS and without VSL iv An already calibrated and validated VISSIM model for Simplified Dynamic Lane Merge System (SDLMS) in accordance with the real life work zone was modified with a VSL through Vehicle Actuated Programming (VAP) code. Three different logics were coded each for VSL alone, early SDLMS+VSL and late SDLMS+VSL. All these logics were fine tuned with several test runs before finalizing it for the final simulation. It is found through the simulation of above mentioned scenarios that for low and medium volume levels (V0500, V1000 and V1500), there is no significant difference between the Maintenance of Traffic (MOT) plans for mean throughputs. However, for higher volume levels (V2000 and V2500), late SDLMS with and without VSL produced higher mean throughputs for all compliance rates and truck percentages except when the demand volume was 2,500 vph and compliance of 60%, where it produces the significantly lower mean throughputs. In terms of travel time through the work zone, results indicated that there is no significant difference between MOT types for demand levels of V0500 and V1000 when compliance is 40% or less but for compliance of 60% and more, only demand volume level that is not significantly different from other MOT types is V0500. This study revealed that VSL increases travel time through the work zone. This might be due to non-compliant vehicles that follow the compliant vehicle v ahead unless they find a sufficient gap in adjacent lane to pass the compliant vehicle. It is also found out that VSL makes the system safer at higher volumes (2,000 vph and 2,500 vph). This was observed through safety surrogate measures selected for this study. Another outcome of this study is that the addition of VSL to the dynamic merge systems helps in improving the overall safety of the system by lowering speed variances and deceleration means of the vehicles travelling through the work zone. The passage of traffic through the work zone is made safer when a speed control is integrated to a dynamic merge system. It can be inferred from the simulation results that integrated SDLMS and VSL systems have better performance in terms of traffic mobility and safety than existing individual controls and also show that the integrated SDLMS and VSL system has more potential than each individual systems.

Improving the predictability of airline schedules in the National Airspace System (NAS) has been a constant endeavor, particularly as system delays grow with ever-increasing demand. Airline schedules need to be resistant to perturbations in the system including Ground Delay Programs (GDPs) and inclement weather. The strategic search heuristic proposed in this dissertation significantly improves airline schedule reliability by assigning airport departure and arrival slots to each flight in the schedule across a network of airports. This is performed using a multi-objective optimization approach that is primarily based on historical flight and taxi times but also includes certain airline, airport, and FAA priorities. The intent of this algorithm is to produce a more reliable, robust schedule that operates in today's environment as well as tomorrow's 4-Dimensional Trajectory Controlled system as described the FAA's Next Generation ATM system (NextGen). This novel airline schedule optimization approach is implemented using a multi-objective evolutionary algorithm which is capable of incorporating limited airport capacities. The core of the fitness function is an extensive database of historic operating times for flight and ground operations collected over a two year period based on ASDI and BTS data. Empirical distributions based on this data reflect the probability that flights encounter various flight and taxi times. The fitness function also adds the ability to define priorities for certain flights based on aircraft size, flight time, and airline usage. The algorithm is applied to airline schedules for two primary US airports: Chicago O'Hare and Atlanta Hartsfield-Jackson. The effects of this multi-objective schedule optimization are evaluated in a variety of scenarios including periods of high, medium, and low demand. The schedules generated by the optimization algorithm were evaluated using a simple queuing simulation model implemented in AnyLogic. The scenarios were simulated in AnyLogic using two basic setups: (1) using modes of flight and taxi times that reflect highly predictable 4-Dimensional Trajectory Control operations and (2) using full distributions of flight and taxi times reflecting current day operations. The simulation analysis showed significant improvements in reliability as measured by the mean square difference (MSD) of filed versus simulated flight arrival and departure times. Arrivals showed the most consistent improvements of up to 80% in on-time performance (OTP). Departures showed reduced overall improvements, particularly when the optimization was performed without the consideration of airport capacity. The 4-Dimensional Trajectory Control environment more than doubled the on-time performance of departures over the current day, more chaotic scenarios. This research shows that airline schedule reliability can be significantly improved over a network of airports using historical flight and taxi time data. It also provides for a mechanism to prioritize flights based on various airline, airport, and ATC goals. The algorithm is shown to work in today's environment as well as tomorrow's NextGen 4-Dimensional Trajectory Control setup.