Coordinated Control of Multirotors for Suspended Load Transportation and Fixed-Wing Net Recovery
MetadataVis full innførsel
As the core technology behind Unmanned Aerial Vehicles (UAVs) have matured, there have been a surge of increased interest for these vehicles over the past years. UAVs have the potential to become a ubiquitous tool for surveillance, remote sensing, inspection, asset and sea monitoring, search and rescue and many other areas. Different types of UAVs are used; fixed-wing UAVs looks like an aircraft, and have long range and high endurance, usually measured in hours. They do, however, usually require special equipment and areas to perform takeoff and landing. Unlike fixed-wing UAVs, multirotor UAVs can produce lift straight upwards, and are thus able to hover in-air, and takeoff and land practically anywhere. Nevertheless, as multirotors must produce all lift from its motors and propellers, they have severely less range, with a typical flight-time of 30 minutes. Multirotor UAVs are mostly used for short-range operations which benefit from their ability to hover. Typical use-cases are inspection, aerial photography and situational overview and surveillance of smaller areas. These are all examples of using the UAV as a remote sensing platform, where the on-board control system ensures the integrity of the UAV alone. This monograph is motivated by the desire to enable multirotor UAVs to interact with the environment in the form of a suspended payload. This is useful for package delivery applications, sensor placement in remote areas such as on icebergs, and autonomous mine-localization. Both single and multiple UAVs cooperatively transporting the same load for increased endurance are considered. In addition, a novel recovery concept for fixed-wing UAVs are presented. By suspending a net below two cooperating multirotors, fixed-wing UAV operations can be performed in situations where landing was previously a challenge, such as ship-based operations and confined rural or forested areas. Chapter 2 gives an overview of the platform developed for the experimental trials conducted in this work. The chapter gives an introduction to the various hardware components and software tools used. The multirotor platform is designed for outdoor usage, and is a hexacopter type platform weighing 2.2 kg, equipped with an autopilot, an on-board computer, precise satellite navigation systems and sensors specifically designed for payload transportation. Chapter 3 considers a single multirotor carrying a suspended load. After giving an overview of related work, dynamic modeling of the interconnected multirotorsuspended load system is performed using Kane’s equations. Based on this model, a nonlinear controller is designed, to guarantee trajectory tracking of the multirotor UAV while being subjected to disturbances from the motion of the suspended load. To minimize the swing of the load, an open loop and a closed loop approach is considered, and combined to achieve robust swing reduction. The proposed controller is verified by numerical simulations and experimental trials. Chapter 4 introduces multi-lift operations, which is the topic for the rest of the monograph. After an overview of the related work, an introduction to modeling and the complex dynamic coupling that occurs when multiple UAVs are transporting the same payload is given. It is illustrated how any motion of any of the UAVs has a direct influence on the motion of the other UAVs in the operation. Equations of motion are derived, which are used for simulation purposes for the rest of the thesis. Chapter 5 presents a kinematic motion controller for multiple multirotors transporting a suspended load. By measuring the relative angles to the suspension point on each UAV, a distributed controller is designed which does not rely on relative position measurements. The controller relies on measuring the weight of the suspended load. Chapter 6 solves the multi-agent control problem by synchronized path-following. A controller is designed on each UAV, which ensures local path-following along a predefined parameterized path. Further, the positions along that path is synchronized, ensuring stability of the desired relative formation. Numerical simulation based on the dynamic models presented in Chapter 4 validates the design. Chapter 7 develops a position based cooperative controller to control the relative position of multiple multirotor UAVs transporting a suspended load. The disturbance from the suspended payload is considered unknown, and an adaptive term is designed to counteract the disturbance on the multirotors, as well as from wind. Numerical simulations and experimental trials verify the results. Chapter 8 presents a novel method for recovery of a fixed-wing UAV in the absence of stationary nets or runways, by suspending a net between two synchronized autonomous multirotor UAVs. This method is especially suited for ship-based operations, where the landing pattern can be optimized with respect to wind direction and speed, and allows for fixed-wing UAV operations to be conducted on smaller vessels without the deck-capacity for large, stationary net. The proposed method is experimentally validated on a small-scale UAV platform. Chapter 9 concludes the thesis and gives some remarks on future work.