TRANSIT: Towards a Robust Airport Decision Support System for Intelligent Taxiing

目前迫切需要更好地利用现有的航空基础设施，因为预计到2035年，空中交通量将增加1.5倍。许多机场的容量接近最大，欧盟委员会认识到有必要增加容量以满足需求。此外，低效率的运营导致延误、拥堵、燃料成本增加和噪音水平，影响到包括机场、航空公司、乘客和当地居民在内的所有利益攸关方。一个关键的问题是飞机地面行动的路线和安排。虽然地面运动只是整个飞行的一小部分，但飞机发动机在滑行速度下的低效率运行可能会导致大量的燃油消耗。这一点尤其适用于大型机场，因为那里的地面机动更加复杂，但也适用于短途飞行，因为滑行在整个飞行中占很大比例。据估计，仅在滑行过程中燃烧的燃料就占短途航班燃料消耗的6%，导致全球每年燃烧500万吨燃料。本计画旨在研究一个决策支援系统，考虑多个因素，以提供更稳健的滑行路线。目前的决策支持系统的路由和调度滑行飞机遭受几个限制：1）他们考虑的唯一目标是最大限度地减少滑行时间，忽略了其他重要因素。这些其他因素包括考虑与燃料消耗、环境影响和成本相关的发动机性能。航线和时刻表，这是有效的燃料和成本，因此，作为一个考虑一维目标的结果妥协。2）机身动力学不考虑在规划的路线和时刻表。因此，发出的滑行指令可能难以遵循，使得遵守分配的路线不现实。3）滑行时间通常基于飞机的平均速度。这是一种过于简单化的说法，这意味着任何超出预期持续时间的滑行动作福尔斯都可能影响其他飞机的滑行。此外，如果采用过于保守的时间缓冲来吸收不确定性的方法，那么由此产生的整体机场运营效率将下降。4）很难为路由和调度系统指定滑行速度和启发式规则，因为：它们取决于机场布局和运营要求，这些要求可以根据空中交通量在一天中变化。因此，路由和调度系统必须重新配置特定的机场和运营约束。5）由于滑行速度的变化和过于简单的飞机模型，缺乏对现实世界中自动路由和调度可以实现多少好处的理解。过境运输系统将直接解决现有系统的这些局限性，更好地利用现有的机场基础设施，减少不断增长的航空部门对环境的影响。多目标优化算法将与飞机模型相结合，以平衡滑行时间、成本和排放的减少。我们的目标是使路由和调度系统很容易重新配置到任何机场。不确定性将直接纳入规划中，从而实现稳健的滑行，并通过飞行员在回路中的试验进行验证。TRANSIT旨在研究这样一个系统及其相关优势，并在广泛的学科和领域（工程，运筹学和计算机科学）之间进行合作，以解决这一具有挑战性的问题。与领先的工业利益相关者的合作以及与知名学者的咨询，确保工作具有前沿性，同时反映工业合作伙伴的需求。

There is an imminent need to make better use of existing aviation infrastructure as air traffic is predicted to increase 1.5 times by 2035. Many airports operate at near maximum capacity, and the European Commission recognises the necessity to increase capacity to satisfy demand. In addition, inefficient operations lead to delays, congestion, and increased fuel costs and noise levels inconveniencing all stakeholders, including airports, airlines, passengers and local residents. A critical issue is routing and scheduling the ground movements of aircraft. Although ground movement is only a small fraction of the overall flight, the inefficient operation of aircraft engines at taxiing speed can account for a significant fuel burn. This applies particularly at larger airports, where ground manoeuvres are more complex, but also for short-haul operations, where taxiing represents a larger fraction of an overall flight. It is estimated that fuel burnt during taxiing alone represents up to 6% of fuel consumption for short-haul flights resulting in 5m tonnes of fuel burnt per year globally. This project aims to investigate a decision support system which considers multiple factors to provide more robust taxiing routes. Current decision support systems for routing and scheduling taxiing aircraft suffer from several limitations:1) The only objective they consider is minimising taxi time, ignoring other important factors. These other factors include taking into account engine performance which is linked to fuel consumption, environmental impact and cost. Routes and schedules, which are efficient in terms of fuel and cost, are therefore compromised as a result of considering a one dimensional objective.2) Airframe dynamics are not taken into account during planning of routes and schedules. Consequently, the taxing instructions issued may be hard to follow, making compliance with the allocated routes unrealistic.3) Taxi time is typically based on average speeds of aircraft. This is an over-simplification meaning that any taxiing manoeuvre which falls outside the expected duration can affect the taxiing of other aircraft. Furthermore, if the approach of including overly conservative time buffers to absorb uncertainty is adopted, the resulting overall airport operating efficiency will be degraded.4) It is difficult to specify taxiing speeds and heuristic rules for routing and scheduling systems as: they depend on airport layout and operational requirements, which can vary throughout the day according to the volume of air traffic. Consequently, routing and scheduling systems have to be reconfigured for specific airports and operational constraints.5) Due to variability in taxi speed and over-simplistic models of aircraft, there is lack of understanding as to how much benefit can be achieved by automated routing and scheduling in real-world settings. TRANSIT will directly address these limitations of current systems, to make better use of existing airport infrastructure and lessen the impact of the growing aviation sector on the environment. Multi-objective optimisation algorithms will be integrated with models of aircraft to balance the reduction of taxi time, cost and emissions. We aim to make the routing and scheduling system easily reconfigurable to any airport. The uncertainty will be directly incorporated in planning, resulting in robust taxiing, verified by pilot-in-the-loop trials.TRANSIT aims to investigate such a system and its associated benefits in collaboration across a broad range of disciplines and fields (Engineering, Operational Research, and Computer Science) needed to tackle such challenging problem. Cooperation with leading industrial stakeholders, and consultation with established academics, ensure that the work is cutting edge while reflecting needs of the industrial partners.