Search

Aircraft trajectory prediction is a feature at the core of most Air Traffic Management (ATM) applications. This process is strongly affected by stochastic sources that impact the accuracy of the predictions. The work presented hereafter focuses on application of the Polynomial Chaos theory to the quantification of the estimated time of arrival at designated waypoints throughout the trajectory. This methodology returns very accurate polynomial representations of trajectory uncertainties with very low computational requirements. This highly efficient approach enables the capability of quantifying the trajectory prediction uncertainties of a real traffic sample (i.e., thousands of flights) within seconds, which is fully compliant with the requirements of future advanced automation tools envisioned in future Trajectory Based Operations (TBO) environment.

In the aspect of Trajectory Based Operations (TBO), accurate trajectory predictions are essential for meeting future concepts defined in SESAR and NextGen visions. The 4D trajectories are results of generated flight plans by Airline Operation Centers to be executed by pilots and monitored by Air Traffic Controllers. Trajectory generation is a process with many input sources such as flight route, speed schedule, initial conditions and environmental conditions. However, the uncertainty in these states generates discrepancies between planned and actual trajectories. This paper distinctly investigates the wind uncertainty by analyzing the forecast data in airline’s operational flight plans and actual wind computed by aircraft’s onboard systems (i.e., Quick Access Recorder). The study goes through the methodology of speed and altitude optimization for cruise phase. The simulation studies reveal the impact of wind uncertainty and potential fuel saving benefits when the precise wind data is available. The case studies are based on historical planned and flown flights.

This paper applies machine learning techniques to improve flight efficiency. Specifically, we focus on two distinct problems: uncertainties in aircraft performance models and uncertainties in wind. In this sense, this paper proposed methodologies to improve baseline models for fuel flow and wind estimations are via operational data. We utilize Base of Aircraft Data (BADA) 4 as baseline for aircraft performance model. Historical Global Forecast System (GFS) predictions are utilized as baseline estimations for $u$ and $v$ components of wind. As for the operational data, Quick Access Recorder (QAR) trajectory footprints of a narrow body and a wide body aircraft, which include actual recorded fuel flow from engines and measured wind speed and direction, are used. State-of-the-art deep learning algorithms are deployed to map baseline estimations for fuel flow and wind to their ground truths. Proper input parameters to have the best estimation results and be compatible with the ground-based flight planning systems are derived through extensive feature engineering. Comparison of the aircraft performance models with real flight data shows that precise estimation of fuel flow with mean absolute errors on a range of %0.1 - %0.7 can be achieved across all the flight modes. Results also show that we can achieve considerable reduction in wind uncertainty both from a mean error and variance sense. For short haul flights, the standard deviations of forecast errors in u and v components are reduced from 6.25 and 8.38 knots to 1.37 and 1.81 knots, respectively. The same reduction is from 11.02 and 10.89 knots to 4.88 and 4.76 knots in the long haul flights.

This study implements deep learning techniques to estimate fuel burn of a jet aircraft. Current ground-based flight planning systems utilize aircraft type specific performance tables to determine fuel flows for given flight conditions and parameters such as altitude, mass and speed. These tables are corrected by a performance factor as the aircraft ages. Despite this update, planned fuel consumption may indeed not overlap with the actual one. In order to synchronize the base aircraft model with aircraft's actual performance, we propose using state-of-the-art deep learning algorithms for building data-driven models of fuel flows. Towards this goal, aircraft's on-board recorded trajectory and parameter data, namely Quick Access Recorder (QAR) data are utilized. The total dataset used within this study comprises of more than 1000 B777-300ER flights from a major European flag carrier airline. The deep neural network architecture is utilized for modeling the actual fuel flow specific to each aircraft and for each major flight mode (climb, cruise and descent namely). We have developed three neural network architectures (according to in-flight and ground based planning use cases) to present a tail-number specific correction factor to Base of Aircraft Data (BADA) models. First architecture involves a QAR data based black-box fuel flow model utilizing in-flight throttle data from all the engines. Comparison of this model with real flight data shows that precise estimation of fuel flow with mean errors lower than %0.1 can be achieved. The second architecture utilizes a physically consistent data regeneration of $\delta_{T}$ (delta thrust) using BADA formulation as to account for the ground planning phase where throttle information is not available. The third model involves a cascaded architecture which utilizes a neural network throttle estimator and the black-box QAR fuel flow model for again the ground planning phase. Comparison of the latter models with real flight data shows that precise estimation of fuel flow with mean absolute errors lower than %0.7 can be achieved at all the flight modes. Initial tests reflect the fact that even better accuracy can be achieved for all models as the data set size increases. Finally, ground based planning fuel flow models are applied to actual flight plans generated by ground based systems. Total trip fuel comparisons show discrepancies up to %3.5 total fuel loading weight, which may result in potential fuel savings by decreasing the fuel load during take-off. Comparison of the planned and the estimated fuel boarding weights (following the actual filed flight plans) for 91 long-haul flights show that fuel burn savings of around 800 [k g] per flight could have been achieved by the proposed methodology at the ground planning phase. For a typical operation of 100 long-haul flights per day, this represents yearly savings on the order of around 17 million USD at current jet fuel prices. This “tail-number specific” performance modeling approach is projected to open considerable frontiers including the in-flight update of performance models through machine learning methods.

This paper presents a physics-guided deep neural network framework to estimate fuel consumption of an aircraft. The framework aims to improve data-driven models’ consistency in flight regimes that are not covered by data. In particular, we guide the neural network with the equations that represent fuel flow dynamics. In addition to the empirical error, we embed this physical knowledge as several extra loss terms. Results show that our proposed model accomplishes correct predictions on the labeled test set, as well as assuring physical consistency in unseen flight regimes. The results indicate that our model, while being applicable to the aircraft’s complete flight envelope, yields lower fuel consumption error measures compared to the model-based approaches and other supervised learning techniques utilizing the same training data sets. In addition, our deep learning model produces fuel consumption trends similar to the BADA4 aircraft performance model, which is widely utilized in real-world operations, in unseen and untrained flight regimes. In contrast, the other supervised learning techniques fail to produce meaningful results. Overall, the proposed methodology enhances the explainability of data-driven models without deteriorating accuracy.

Considering the transformation in roles of existing air traffic management technologies, future flight operations and flight deck systems will need additional avionics and operational procedures that involve adaptive algorithms and advanced decision support tools. The main purpose of this article is to provide a theoretical framework for tactical 4D-trajectory planning and conflict resolution of an aircraft equipped with novel automation tools. The proposed 4D-trajectory-planning method uses recent algorithmic advances in both probabilistic and deterministic methods to fully benefit from both approaches. We have constructed an aircraft performance model based on Base of Aircraft Data 4 with high-level hybrid flight template automatons and low-level flight maneuver automatons. This multi-modal flight trajectory approach is utilized to generate cost-efficient local trajectory segments instead of solving complex trajectory-generation problems globally. The proposed sampling-based trajectory planning algorithm spatially explores the airspace and provides proper separation through local trajectory segments and guarantees asymptotic optimality under certain conditions. Moreover, we have integrated the cross-entropy method, which transforms the sampling problem into a stochastic optimization problem, rapidly converges on the minimum cost trajectory sequence by utilizing available flight plans, and reduces the amount of sampling. The integration of the proposed strategies lets us solve challenging, real-time in-tactical 4D-trajectory planning problems within the current and the envisioned future realm of air traffic management systems.

Aircraft trajectory prediction refers to an estimate of the future positions of a flight for a given look ahead time. In the context of achieving safe and seamless traffic management, efficient and modular trajectory prediction tools will be the crucial functions in future trajectory based operations. The main objective of this paper is to present a highly scalable trajectory generation infrastructure (TGI) that is capable of predicting 4D trajectories for given initial conditions and flight plans, detecting conflicts and providing separation maneuvers, avoiding restricted regions declared as NOTAM, calculating air sector entry/exit and occupancy times. The proposed TGI includes advanced aircraft performance models based on BADA (Base of Aircraft Data), standard navigation functions, standardized interface for operational rules of airspace inputs (i.e. airports, navaids, fixes, Standard Instrument Departure Route (SID), Standard Arrival Route (STAR) procedures), trajectory control algorithms, conflict detection functions and conflict resolution action library. The developed tool provides a comprehensive platform for air traffic control research, and the its functions are explained throughout this paper.

The increasing demand in the air transportation has been bringing about increased workload to air traffic controllers. Reducing the workload, hence increasing the airspace capacity could be enabled by developing automated air traffic management tools. Our previous work presented a new hybrid system description, namely automated AT Co, modeling the decision process of the air traffic controllers in en-route and approach operations. The developed tool also considers enhanced air traffic and aircraft dynamics. The hybrid system provides realistic conflict resolution maneuvers in 3D space in reasonable computation times. The trajectory prediction infrastructure behind the developed tool accepts mainly flight plans and aircraft performance variables (i.e. initial conditions, performance model) as inputs to yield trajectories. However, some aircraft specific parameters are not exactly known for ground based systems. These can be described as random variables. This phenomena results in uncertainties in trajectory prediction. In this paper, trajectory predictions during climb phase are improved through model driven state estimation. The algorithm uses observed track of an aircraft obtained from a period of time and recovers the take-off mass error considering the conservation of energy rates. It is shown that trajectories are improved in both in time and spatial terms compared to predictions with nominal states.

Efficient trajectory prediction tools will be the crucial functions in future trajectory-based operations (TBO). In addition to win and controller actions, uncertainties in climbing flights are major components of prediction errors in a flight trajectory. Due to the operational concerns, aircraft take-off weight and climb speed intent, which are key performance parameters that define climb profiles, is not entirely available to round-based trajectory prediction infrastructure. In the scope of air traffic flow management, sector entry and exit times, including where the climb ends and descending starts, are the main inputs for demand- capacity balancing processes. In this work, we have focused on uncertainties over climb trajectory to quantify and analyze their impact on climb times to cruise altitudes. We have used model-driven data statistical approaches through aircraft flight record data sets (i.e. QAR). As result of this analyze, probabilistic definitions are generated for aircraft take-off weight and speed intent. The regression between these climb parameters and flight distance is acquired to reduce the uncertainty at strategical level. Moreover, reducing climb uncertainty through adaptive uncertainty reduction is also demonstrated at the tactical level of flight. Through the simulations, the impact of reducing the uncertainty in aircraft mass on climb time is illustrated.

The future flight deck will need additional avionics and operational procedures which involve adaptive algorithms and advanced decision support tools, considering the transformation of redefining existing roles in ATM technologies. These tools should also meet the requirements of the future flight operations defined in NextGen and SESAR 2020+ visions. Following this line of thought, the main purpose is to develop a theoretical framework for both in-tactical 4D trajectory planning and collision avoidance problems of an aircraft equipped with automated tools, which in turn can almost achieve free flight. The proposed approach in 4D trajectory planning involves recent algorithmic advances in both probabilistic and deterministic methods, in order to fully benefit from the powerful and useful capabilities of both approaches. Furthermore, we have constructed an aircraft performance model based on BADA 4 so as to generate trajectories. The multi-modal approach for the aircraft model is utilized to overcome the problem of the complex optimal trajectory generation via reducing the dimension. The sampling-based algorithm embeds this local trajectory generation procedure and guarantees asymptotic optimality under certain conditions. Moreover, a cross-entropy method has been integrated to enable highly efficient sampling. We have also formulated the collision avoidance problem as a perfect-information zero-sum differential game, as well as introduced a sampling-based procedure integrating sampling-based motion planning algorithms to generate feasible approximate solutions. Finally, different scenarios have been simulated, and pilot interactions with the integrated B737-800 Flight Deck Testbed deployed in ITU ARC have been demonstrated.