Search

En el trabajo de esta tesis doctoral se proponen ciertas mejoras para el Air Traffic Management (ATM) en el área del aeropuerto. Concretamente se ha propuesto una serie de mejoras para el sistema Advanced Surface Movement Guidance and Control System (A-SMGCS) que es el encargado de las funciones de enrutado, guiado, vigilancia y control de las aeronaves en dicha área. Es en estas dos últimas funciones en las que se ha centrado el trabajo. La función de vigilancia se encarga principalmente de, a partir de las medidas de los sensores disponibles, obtener con precisión la posición de los blancos, así como su identificación. Para ello, esta función tiene que realizar las tareas de asociación de las medidas, corrección de sesgos y filtrado de las medidas. Se propone una arquitectura de corrección de sesgos basada en la división de éstos en tres, según la fuente que los genera. En concreto se estima el sesgo dependiente solo del sensor, el sesgo dependiente de la pareja blancosensores y por último el offset que existe en las medidas verticales según el sensor sea geométrico o barométrico. Respecto al filtrado de las medidas, se propone un sistema que utiliza dos filtros en paralelo con una lógica de selección para obtener la estimación de la posición de los blancos. El primer filtro utilizado es uno que utiliza la información contenida en el mapa del aeropuerto para mejorar las prestaciones. Para ello, este filtro está basado en un enfoque bayesiano de múltiples hipótesis para determinar las carreteras por las que se mueve el blanco. Además, el sistema descompone el movimiento del blanco en su componente longitudinal y transversal. El segundo filtro que compone el sistema es un filtro robusto que no utiliza la información del mapa y por tanto es capaz de obtener la estimación en situaciones en las que el filtro basado en el mapa no es capaz (aviones aterrizando, salidas de pista, etc.). Además, tanto el sistema de seguimiento completo como cada filtro por separado han sido evaluados y sus prestaciones han sido comparadas con otros filtros de literatura. De esta manera se ha podido demostrar que el sistema de seguimiento es competitivo y presenta una mejor calidad respecto a otras soluciones. La función de control se encarga de detectar los posibles conflictos que pueda haber en el área del aeropuerto. Utiliza las posiciones filtradas de los blancos, obtenidas por la función de vigilancia, así como las rutas establecidas, obtenidas por la función de guiado. En esta tesis se ha diseñado y analizado una safety net capaz de detectar los siguientes conflictos: incumplir la separación horizontal mínima entre aeronaves, exceder la velocidad máxima de las taxiways y las incursiones en las runways. Las safety nets diseñadas se basan en una caracterización rigurosa del problema y pueden funcionar o sólo con las posiciones de los blancos o utilizando también la información del mapa, mejorando así las prestaciones. Adicionalmente, se ha medido la calidad de las safety nets desarrolladas, demostrando su correcto funcionamiento así como la reducción drástica de probabilidad de falsa alarma y detección anticipada de los conflictos cuando se usa la información del mapa. Por último, se presentan las principales conclusiones de esta tesis, así como las posibles futuras líneas de investigación. ----------ABSTRACT---------- In this Ph.D thesis, certain improvements for the Air Traffic Management (ATM) on the airport area are proposed. Specifically, a series of improvements has been proposed for the Advanced Surface Movement Guidance and Control System (A-SMGCS) system which is in charge of the routing, guidance, surveillance, and aircraft control functions in that area. The work done has been focused on these last two functions. The surveillance function is mainly in charge of, from the available sensors measures, obtaining accurately the position of the targets as well as their identification. To do this, this function has to perform the following tasks: association of measures, correction of biases, and filtering of measures. A bias correction architecture is proposed based on the division of them according to the source that generates them. To be more specific, the biases estimated are: the bias dependent only on the sensor, the bias dependent on the targetsensors pair, and finally, the offset that exists in the vertical measurements according to whether the sensor is geometric or barometric. Regarding the filtering of the measurements, a system that uses two filters in parallel with a selection logic to obtain the estimation of the position of the targets is proposed. The first filter uses the information contained in the airport map to improve the performance. This filter is based on a Bayesian Multiple Hypotheses approach in order to determine the segments being followed by the target after road junctions. In addition, this filter decouples the target movement in two components, one related to along road movement, and another one designed to track movement transversal to road. The second filter of the tracking system is a robust filter that does not use the map information and is therefore capable of obtaining the estimate in situations where the map-based filter is not (landing aircraft, runway excursions, etc.). In addition, both the complete tracking system and each filter have been evaluated and their performance has been compared to other filters in the literature. Hence, it has been possible to demonstrate that the designed tracking system is competitive and presents a better performance than other solutions. The control function is responsible for detecting possible conflicts that may exist on the airport area. It uses the filtered positions of the targets obtained by the surveillance function, as well as the established routes obtained by the guidance function. In this thesis a safety net capable of detecting the following conflicts has been designed and analysed: breaching the minimum horizontal separation between aircraft, exceeding the maximum speed of the taxiways, and the runways incursions. The designed safety nets are based on a rigorous characterization of the problem and can work either with only the filtered positions of the targets or also using the map information, thus improving the benefits. In addition, the designed safety nets have been evaluated, demonstrating their correct performance as well as the drastic reduction of the false alarm probability and the early detection of conflicts when the information on the map is being used. Finally, the main conclusions of this thesis are presented as well as the possible future lines of research.

Small and medium-sized airports are poorly occupied or occupied for a small period of time, hence the remote control towers are a good solution to control air traffic in these airports. A surveillance and monitoring air traffic prototype has been built to test surveillance, control and visualization concepts. The prototype is comprised of a hybrid real-synthetic scenario using augmented reality techniques to evaluate the operation of the surveillance/vision systems, observing the behavior of this system with different targets, backgrounds, inclement weather, etc. and helping the design/experimentation of novel control procedures in the airport area. This allows the analysis of risky situations and the controller training without putting at risk neither people nor material goods.

In this paper, a complete and rigorous mathematical model for secondary surveillance radar systematic errors (biases) is developed. The model takes into account the physical effects systematically affecting the measurement processes. The azimuth biases are calculated from the physical error of the antenna calibration and the errors of the angle determination dispositive. Distance bias is calculated from the delay of the signal produced by the refractivity index of the atmosphere, and from clock errors, while the altitude bias is calculated taking into account the atmosphere conditions (pressure and temperature). It will be shown, using simulated and real data, that adapting a classical bias estimation process to use the complete parametrized model results in improved accuracy in the bias estimation.