No abstract

These days we demand Unmanned Aerial Systems (UASs) fly autonomously and be able to physically interact with their environment executing prescribed tasks. An excellent example of that are the Aerial Manipulators (AMs) i.e. UASs formed by the join of an Unmanned Aerial Vehicle (UAV) and a Robot Manipulator (RM). Moreover, the lack of structured workspaces in outdoor operations is challenging for the control system, forcing to increase notably its complexity to meet such requirements but keeping in mind the trade-off between the task performance and computational burden. In this work, a nonlinear control strategy is proposed and thoroughly tested on an AM. The strategy combines the use of robust controllers separately for both UAV and RM exploiting their stability margins to optimise different prescribed criteria in real time. The inclusion of this optimisation in the loop shows excellent results, sharing priorities of the controllers as required. Following this idea, two different strategies have been tested in a benchmark system showing promising results and, furthermore, feasible for a subsequent implementation in the available platform.

In the last years, the research on unmanned aerial systems (UASs) has shown a marked growth and the models to simulate UASs have been deeply studied. Although onboard controller algorithms have increased their complexity, most of them still rely on simplistic models. In essence, aerodynamic forces/torques are generally considered either insignificant compared to propulsion and inertial forces or acceptably modeled with constant aerodynamic coefficients estimated in a particular flight regime. However, the increase of power in the onboard computers allows to make controller algorithms more complex, and therefore, to increase the total performance of the UAS. In this regard, this work provides an explicit aerodynamic model for multirotor UAS that, unlike most of the current models, does not need iterations to be adjusted to the flight conditions at a higher computational cost. This explicit nature makes it an excellent choice for being implemented in onboard computers, thus covering a broad range of applications, from controller design to numerical analysis (e.g., the capture nonlinear phenomena like bifurcations). To obtain this accurate explicit mathematical aerodynamic model, a thorough analysis of a batch of simulations is carried out. In these simulations, the aerodynamic forces and torques are estimated using computer fluid dynamics (CFD), and the propulsive effects are taken into account via blade element momentum theory (BEMT). A study of its implementation for different regimes and platforms is also provided, as well as some potential applications of the solution, like robust control strategies or machine learning.

Recently, the complexity of control systems for autonomous Aerial Manipulators (AMs), i.e. Unmanned Aerial Vehicle (UAV) + Robot Manipulator (RM), is growing faster as per our demand of being able to perform more and more complex tasks. In the present work, we go a step forward adding an optimiser to the actual (nonlinear) control strategy, in order to comply with high-level control demands related to safety, accuracy and efficiency of the operational task. The actual strategy combined robust controllers with separate referencesfor both aerial vehicle and robot manipulator-optimising their shared priorities. Here, we demand the controller to meet an additional feed-forward action, so that the a priori free degree of freedom of the UAV relative-pose reference is optimised in real time according to the aforementioned requirements. In particular, the influence of the UAV relative-pose reference on the capabilities of the AM has been thoroughly analysed, demonstrating among others the benefits of a correct configuration to meet such high-level requirements, while reducing the End-Effector (EE) error, preventing unstability of hazardous situations and increasing the energetic efficiency of the whole system. A complete analysis of realistic simulations on a benchmark AM is reported.

In this paper, an adaptive nonlinear strategy for the motion and force control of flexible manipulators is proposed. The approach provides robust motion control until contact is detected when force control is then available-without any control switch-, and vice versa. This self-tuning in mixed contact/non-contact scenarios is possible thanks to the unified formulation of force and motion control, including an integral transpose-based inverse kinematics and adaptive-update laws for the flexible manipulator link and contact stiffnesses. Global boundedness of all signals and asymptotic stability of force and position are guaranteed through Lyapunov analysis. The control strategy and its implementation has been validated using a low-cost basic microcontroller and a manipulator with 3 flexible joints and 4 actuators. Complete experimental results are provided in a realistic mixed contact scenario, demonstrating very-low computational demand with inexpensive force sensors.