Causal models for performance evaluation of added-value operations

Abstract

La presente tesis doctoral se presenta en forma de compendio de publicaciones, en la cual se han desarrollado diversos Modelos Causales para ayudar en la toma de decisiones usando como herramienta principal las relaciones causa-efecto inherente del sistema. A continuación se describen brevemente los artículos incluidos en la tesis doctoral El artículo titulado “A CD&CR causal model based on path shortening/path stretching techniques” publicado por la revista Transportation Research Part C: Emerging Technologies esta enfocado hacia la investigación académica de alto nivel que aborda el desarrollo, las aplicaciones y consecuencias, de las nuevas tecnologías en el campo del transporte. El enfoque del modelo causal desarrollado en Redes de Petri Coloreadas aportan no solo una metodología innovativa per se, sino que su aplicación e implicaciones en la etapa de planificación, diseño, y gestión, es de particular interés en el rendimiento general del sistema de transporte aéreo. En artículo titulado "Revisiting the pallet loading problem using a discrete event system approach to minimise logistic costs” publicado en la revista International Journal of Production Research (IJPR) se presenta un modelo orientado a eventos discretos para abordar el Problema del Palletizado (PLP) usando como formalismo de especificación las Redes de Petri Coloreadas. El problema de Paletizado conforma una parte fundamental en el flujo logístico de diversos sistemas de transporte (terrestre o aéreo), siendo un factor clave en etapas aguas arriba (e.g. producción) y abajo del sistema (e.g. distribución). El trabajo presenta una novedosa forma de modelado con una importante mejora de los kPI’s respecto a los actuales entornos comerciales, soportando tanto diferentes heurísticas así como su integración con el sistema para encontrar configuraciones óptimas para el acomodo de cajas con una tipología diversa. El artículo denominado "Integrating and sequencing flows in terminal maneuvering area by evolutionary algorithms," publicado por la IEEE/AIAA en el Digital Avionics Systems Conference (DASC), 2011 presenta una aproximación innovadora para aliviar el problema de congestión en el espacio aéreo terminal (terminal maneouvering área). En este artículo se analiza un procedimiento alternativo para la integración y secuenciación de las aeronaves en la fase de aterrizaje. El artículo ha recibido diversas menciones tal como el mejor artículo de la sesión, mejor artículo en la sección de mejoras a la capacidad del sector aéreo y mejor artículo de estudiante. Estas menciones indican la importancia y trascendencia de dicho trabajo donde se utilizan algoritmos desarrollados para la optimización de sistemas altamente complejos como es el espacio aéreo terminal. El artículo que lleva por nombre “A TMA 4DT CD/CR causal model based in path shortening/path stretching techniques”, publicado en el International Conference on Research in Air Transportation — ICRAT 2010, ha recibido la mención Best paper award reconociendo las aportaciones del trabajo presentado. El modelo desarrollado en Redes de Petri Coloreadas para la solución de conflictos entre aeronaves presenta resultados experimentales validados mediante simuladores de vuelo de la compañía Aireuropa que permiten validar las trayectorias obtenidas mediante el modelo desarrollado. Finalmente, el articulo Causal model to sequence and merge 4DT flows in TMA, se encuentran publicados en las memorias del International Conference on Application and Theory of Automation in Command and Control Systems ATTACS2011 donde se presenta un algoritmo de solución para la secuencializacion y fusión de flujos de aterrizaje de aeronaves donde se muestra el gran potencial para mejorar de la toma de decisiones usando el espacio de estados de los modelos de RdPC.

The present PhD thesis report has been elaborated as a compendium of publications, in which diverse Causal Models have been developed to assist in the decision making process using a cause-effect relationship approach inherent in the system. A brief description of the items included in the doctoral thesis. The document is organized in four different parts. First, the Chapter called “Basic Notions” introduces the basic notions and a general perspective on the systems approach. Particular interest has been placed in the Discrete-Event Systems approach, presenting the main features of this formalism. The main theory behind the Coloured Petri Nets approach is presented in Section 2.1.1 Different case studies are provided from Chapters 3 to 5. Chapter 3 presents the work named “A CD&CR causal model based on path shortening/path stretching techniques”, which has been accepted for publication in the Journal Transportation Research Part C: Emerging Technologies. This work depicts a groundbreaking approach to alleviate the airspace congestion and to deal with the implications for the planning, design, and management of Air Traffic Control operations. Particular interest has been placed Terminal Manoeuvring Area (TMA) where the traffic conditions impose to take time-critical decisions. Therefore, an appropriate management of arrival operations could alleviate congestion which impacts directly in capacity and efficiency of the overall Air Traffic Management operations. To evaluate the performance of the the resolution strategy to avoid non-efficient procedures, diverse scenarios have been tasted in a busy traffic period at Gran Canaria’s airport. The result obtain contribute with the planning and management of Air Traffic Control operations to increase the overall predictability of the Air traffic, with benefit to airlines and airports, among others. Chapter 4 corresponds to the article "Revisiting the pallet loading problem using a discrete event system approach to minimise logistic costs” published in the International Journal of Production Research (IJPR). This work presents an innovative and challenging modelling approach to optimize the space and distribution of boxes into a pallet, supporting the inherent box diversity (heterogeneous palletizing problems) of present production and distribution logistic systems. The space utilization is modelled as squares that can be fragmentise and de-fragmentise. In a first approach, the state space analysis is performed to evaluate different optimal configurations to load the maximum number of boxes on a rectangular pallet. The second approach implements heuristics to show that acceptable occupancy results can be obtained without requiring the exhaustive evaluation of the different feasible combination. Chapter 5 introduces the work "Integrating and sequencing flows in terminal maneuvering area by evolutionary algorithms" in proceeding of the IEEE/AIAA en el Digital Avionics Systems Conference (DASC), 2011. This work has received three mentions: Best student paper award; Best paper in the ATM Capacity Improvements track award; y Best paper of session award which confirm the transcendence and implications of such approach. This chapter presents a new approach to optimize a set of aircraft planned to land at a given airport. It is proposed to merge the incoming flows from different routes by mean of speed and path changes. Those changes aim to remove conflicts at merging points and to maintain separation of aircraft following the same route link according to their wake turbulence constraint. The optimization criteria are based on the minimum deviation from the initial path planning. This algorithm has been successfully applied to Gran Canaria airport in Spain with real traffic demand samples for which conflict free flow merging is produced smoothly with optimal runway feeding. Finally, Chapter 5 contains the overall conclusions, future work, summary of contributions and complementary publications on the author. The work presented in annex called “Causal model to sequence and merge 4DT flows in TMA” summarizes the development of a decision support algorithm to tackle the merging and sequencing problem within the Terminal Manoeuvring Area (TMA) sector. The algorithm works with multiple landing traffic flows that share the airspace in the same time window. A flexible terminal area route structure is proposed; it eliminates conflicts within the Standard Terminal Arrival (STAR) while traffic is merged prior converging on to the final approach. As a first instance the first come first serve (FCFS) sequence policy will be tasted and later on the sequence will be altered by means of the Constrained Position Shifting (CPS) algorithm. Gran Canaria STAR is used to evaluate the benefits of the proposed model under synthetic traffic; and to determine the spacing buffers the ICAO Separation minima (ICAO DOC-4444) criteria is used as on current methodology. Finally, in Annex, the paper “A TMA 4DT CD/CR causal model based in path shortening/path stretching techniques” introduces the basic ideas behind a discrete event model for Conflict Detection and Conflict Resolution algorithm in a TMA 4D trajectory scenario in presented which focuses mainly on the arrival phase. This model brings a very interesting knowledge about the events that take place in the management of 4DT and their interactions in Gran Canaria TMA to remove non-effective operations, avoid delay propagation between arrivals and optimize the occupancy of the runway. The causal model developed considers different alternative predefined turning points for each flight evaluating path shortening/path stretching of all trajectories upwards the merging point in a TMA.