Since the birth of commercial aviation, the applications and benefits of aircraft have grown immensely. This, in perfect synchrony with the average increase of purchasing power of the society, has rocketed the number of aircraft flying the skies. This increase comes at a cost, both in environmental and airspace capacity aspects. This thesis works towards the alleviation of the issues caused by the high number of flights, proposing concepts and mechanisms to safely increase the airspace capacity whilst minimising the environmental impact of aviation. This incredibly complex and neverending pursuit is omnipresent in the literature. One promising topic is the four dimensional (4D) trajectory optimisation with higher levels of automation.
The research in this PhD thesis proposes an integrated framework for trajectory optimisation, trajectory prediction and parameter estimation, with which new air traffic management concepts can be assessed. This framework has the flexibility to optimise trajectories ranging from a free-flight to a very strict route structure, from a complete freedom at the vertical profile to a specific adherence to flight levels, etc. The 4D optimisation strategy results in a trajectory that complies with the scenario characteristics, which minimises a given functional objective such as the operational cost, time, fuel, etc. Furthermore, the same framework is used in a novel strategy to perform adaptive trajectory prediction (with conformance monitoring), and to estimate unknown parameters of an aircraft.
To resolve this problem, an optimal control problem is formulated and converted into a non-linear programming (NLP) problem with direct collocation methods, and numerically resolved by an NLP solver. A comprehensive software architecture is presented, taking benefit from the best of two worlds to enable the flexibility and genericity of the developed optimisation framework: an object-oriented software coding language (C++) and a very powerful algebraic modelling language (GAMS).
Based on this optimisation framework, the thesis produces operationally relevant results, demonstrating that the framework can cope with a variety of problems, and contributing to the ultimate goal of safely increasing airspace capacity and air traffic efficiency. Illustrative examples are presented focussed on the departure phase within a terminal manoeuvring area. First, an assessment of the efficiency of required times of arrival as a ways to increase air traffic capacity is presented, providing results on the cost in terms of fuel and time of imposing these time requirements within a TMA (which can get to surprisingly low figures), and its effectiveness for traffic separation. Second, the implementation of an aircraft separation methodology is presented, where an intruder trajectory is predicted and the ownship calculates its own optimal trajectory that deviates from it. A conformance monitoring strategy is implemented to ensure that the separation is maintained throughout the flight, acknowledging deviations, and reacting accordingly. Third, the prediction of the intruder trajectory is enhanced by the estimation of an equivalent mass using known past states. An impressive accuracy is achieved early after the beginning of the flight. Finally, the implementation of a multi-aircraft separation strategy is presented, where multiple aircraft are simultaneously optimised in the same optimisation problem, all whilst maintaining separation between them. The complexity of the alignment of aircraft coordinates for a fair comparison is tackled from a novel perspective.
Conclusively, the different strategies for aircraft separation are compared, and quite surprisingly the best results for each strategy are quite similar. Indeed, the increase in operational cost that the different strategies present (when compared to the individual optimal trajectory) is negligible and alledgedly better than the current air traffic control separation paradigm.
© 2001-2024 Fundación Dialnet · Todos los derechos reservados