Trajectory Control of Very Flexible Aircraft

在追求减少碳足迹和更可持续的运输方式的过程中，商用飞机正朝着更高效的设计发展。波音787和空客A350等新型机翼采用了复合材料，使机翼重量轻得多，而且能够显著变形。这些大变形要求将气动弹性耦合结构和气动模型集成到飞机设计过程中，以预测机翼大变形、非定常气动力和飞机飞行动力学之间的行为和相互作用。最重要的是要清楚地了解这些相互作用，因为它们可能会造成灾难性的后果，正如我们从过去的事故中看到的那样，比如2003年美国宇航局的太阳神号飞机。此外，功能强大和重量轻的电子元件的发展正在产生高空伪卫星（HAPS）的概念，用于在全球范围内提供数据通信、成像等服务。与轨道飞行器相比，HAPS具有更可控的飞行轨迹和良好的可维护性。这些平台，就像空中客车公司的“西风”一样，旨在通过单独使用太阳能，在远高于天气系统的高海拔地区一次停留数月。在这种程度上，只有设计具有极高的空气动力学效率，实现非常高的展弦比翼轻结构可以考虑。这样的机翼非常纤细，并且在巡航条件下具有与翼展相当的变形，引入了空气动力学和结构动力学之间危险相互作用的可能性，正如太阳神所经历的那样。控制这些飞机是手头的挑战，因为在飞行任务期间车辆几何形状的巨大变化通常需要使用模型预测控制策略，能够随着车辆变形而更新其内部模型。然而，描述飞机的空气动力学，结构变形和飞行动力学具有足够的保真度水平的模型通常在尺寸方面是禁止的，不能直接用于控制的目的。因此，必须求助于模型缩减技术，以减小此类系统的尺寸，使控制器能够在现代硬件中运行。模型简化过程本身并不是微不足道的，因为必须保留最重要的动态，并且底层系统的大小足够大，从而对计算构成挑战。因此，将研究为不同飞行条件下的飞机创建预先计算的降阶模型数据库的概念，以便控制器在它们之间“在线”进行插值。这将允许“离线”执行所有繁重的计算。然后，控制器将保证飞机在其飞行包线内的稳定性，同时优化资源受限的飞行轨迹，因为在24小时周期内可以提取有限的太阳能。该项目将使用和开发开源的内部工具SHARPy（在Python1中模拟高纵横比飞机），该工具具有收集描述这些新高空平台的复杂和大型气动弹性模型所需的能力，执行模型缩减，模型之间的插值，并最终纳入飞行控制方法。这需要采用控制理论和模型简化的最新技术，将其应用于手头的气动弹性问题，同时将计算成本控制在现代硬件所能承受的范围内。这将有利于颠覆性高空平台的设计，这些平台有望提供从偏远地区的互联网连接到野火演变成像等服务，我们可以在中短期内看到这些平台投入使用。

In the pursuit of a reduced carbon footprint and more sustainable transportation methods, commercial aircraft are evolving toward more efficient designs. New wing concepts such as those found in the Boeing 787 and Airbus A350 incorporate composite materials that allow for much lighter wings that are capable deforming significantly. These large deformations require that aeroelastic - coupled structural and aerodynamic - models are integrated into the aircraft design process to predict the behaviour and interaction between large wing deformations, unsteady aerodynamic forces and the aircraft's flight dynamics. It is paramount to have a clear understanding of these interactions since they can have catastrophic consequences, as seen through past mishaps like that of the NASA Helios aircraft in 2003. In addition, developments in powerful and lightweight electronic components are giving rise to the concept of high-altitude pseudo-satellites (HAPS), used to provide data communication, imaging etc. services worldwide. Compared to their orbital counterparts, HAPS have the added benefit of a more controllable flight trajectory and good maintainability prospects. These platforms, like the Airbus Zephyr, aim at staying airborne for months at a time at high altitudes, well above weather systems, by using solar power alone. To such extent, only designs with extreme aerodynamic efficiencies that are achieved by very high aspect ratio wings with lightweight structures can be considered. Such wings are very slender and in cruise conditions have deformations that are comparable to the wing span, introducing the possibility of hazardous interactions between aerodynamics and structural dynamics as experienced by Helios.Controlling these aircraft is the challenge at hand since the large changes in the geometry of the vehicle during the flight mission often require the use of model predictive control strategies capable of updating their internal model as the vehicle deforms. However, models that describe the aircraft's aerodynamics, structural deformations and flight dynamics with an adequate level of fidelity are typically prohibitive in terms of size and cannot be used directly for the purposes of control. Therefore, one must turn to model reduction techniques to reduce the size of such systems to enable controller operation in modern hardware. The model reduction process itself is not trivial since the most important dynamics must be retained and the size of the underlying system is large enough to pose a computational challenge. Consequently, the concept of creating a database of pre-computed reduced order models for the aircraft in different flight conditions for the controller to interpolate between them "online" will be investigated. This will allow for all the heavy computations to be performed "offline". The controller then will guarantee stability of the aircraft within its flight envelope while optimising the resource-constrained flight trajectory given the limited use of solar energy that can be extracted during a 24 hour cycle.This project will use and develop the open-source, in-house tool, SHARPy (Simulation of High Aspect Ratio airplanes in Python1) with the capabilities necessary to gather the complex and large aeroelastic models describing these new high-altitude platforms, perform model reduction, interpolation between models and finally incorporate flight control methods. It will require adopting the latest state-of-the-art techniques from control theory and model reduction to apply them to the aeroelastic problem at hand, all while keeping computational cost to within the reach of modern hardware. This will hopefully benefit the design of disruptive high-altitude platforms envisaged to provide services ranging from internet connection in remote areas to imaging the evolution of wildfires and that we could see entering service in a short to medium time frame.