Analysis, Modelling, and Optimization of Gate-To-Gate Aircraft Operation for Enhanced Air Traffic Management

Abstract

The global air passenger demand will surpass the 10 billion mark by the end of 2040. India itself is poised to become the third largest aviation market and traffic generator by 2033, with a forecast of over 950 million passengers and 10 million metric tons of cargo. This rapid growth in air travel demand raises critical challenges for the existing and future air traffic management (ATM) system. The present ATM system is on the verge of attaining its operational limits, and further demand-capacity imbalances can result in large delays, breakdown of airline schedules, and critical safety breaches. Therefore, the interest of current and future ATM systems lies in technological advancements in airspace management and movement optimization, aiming to support gate-to-gate flight operations. The overarching goal of the Airports Authority of India (AAI), the International Civil Aviation Organization (ICAO), and other aviation agencies also point towards the same direction. Hence, this thesis aims to advance state-of-the-art in ATM strategies for more efficient gate-to-gate aircraft operations in air transportation systems. The thesis focuses on three key components of ATM: terminal airspace (TMA) departure management, TMA arrival management, and en-route airspace management. For better departure and arrival management, we developed a data-driven descriptive framework to measure and assess the TMA traffic flow characteristics using flight record data. The framework involves the identification and extraction of key metrics representing TMA characteristics and performing cross-sectional analysis for different flight types and peak periods. Simultaneously, we proposed another data-driven statistical framework to analyse and model the dynamic evolution, fluctuation characteristics, and time-varying patterns of TMA traffic. The framework utilizes seasonality analysis, multifractal detrended fluctuation analysis, distribution functions and queue modelling. Furthermore, we developed quasi-stochastic mixed integer programming (MIP) models to optimally schedule departures and arrivals based on TMA complexity, TMA traffic, and safety constraints. For better en-route airspace management, we propose an integrated framework of multicommodity flow optimization, unconstrained multi-objective network clustering, and a rule-based approach to partition the airspace sector into sub-sectors to be used for en-route traffic monitoring. The frameworks and models were implemented and validated using Chennai airport, Chennai TMA, and a sector of Indian airspace as case studies. The results suggest that TMA flow metrics vary with aircraft wake, and high variance in metrics on continuous hours reflect congestion & inefficient operations. TMA traffic also exhibited dynamic fluctuation and time-varying stochasticity. MIP models were implemented dynamically through a rolling horizon framework. MIP models generate optimal departure and arrival schedules with minimum operational delay and conflicts. The optimal departure and arrival schedules also ensure no loss of runway throughput. Finally, the integrated sub-sectorization framework generates balanced, convex, and compact sub-sectors, reflecting system-wide demand and network behaviour. Such sub-sectors can lead to better workload balancing and enhanced en-route traffic monitoring capabilities. Research output from this thesis is expected to advance the existing literature and strategies in ATM. Specific contributions include (i) improved TMA performance and congestion assessment strategies, (ii) safe and optimal TMA traffic management solutions, and (iii) a fast adaptive en-route flow control and monitoring platform.