This paper deals with the problem of modelling and controlling an innovative aerial manipulator, i.e. a vertical take-off and landing aircraft equipped with a fully-actuated robotic arm. This system is able to perform complex operations that require the physical interaction with the surrounding environment while remaining airborne. Once a detailed dynamical model in the planar case is provided, a control law able to govern all the degrees of freedom of the system is discussed. Beside the methodological contribution that is easily extendable to different combinations of UAVs and robotic arms, the effectiveness and main properties of the proposed control algorithm are illustrated with the help of an experiment that takes into account the case in which the manipulator is in contact with the surrounding environment.

This paper provides an architectural description from the software point of view of the simulator environment developed for the AIRobots project. The scope of the project is the realization of an aerial service robotic prototype, a sort of robotic hand to be employed in inspection-by-contact tasks. The simulator is then crucial in both the training of the human operator, and as a support tool for the development and validation of low- and high-level control algorithms. The tasks that can be performed are not limited to free-flight missions, but include also to the cases in which the robot has to actively interact with the environment. The simulator relies on Simulink and Blender, and has been designed with a modular structure that makes software-in-the-loop and hardware-in-the-loop simulations possible by simply replacing the different control modules with the real controllers on the prototypes.

This paper focuses on the modeling and control of a flying robot. The complete system, composed of a quadrotor unmanned aerial vehicle and a custom-made manipulator, has been designed for remote inspection by contact of industrial plants. The goal of this paper is to show the dynamical characteristics of the flying robot during tasks that require physical interaction, and to determine a control strategy that allows to safely interact with unknown environments. The methodology has been implemented on a real prototype and tested in an indoor area. Experimental results validate the proposed controller and show its effectiveness.

This video shows a special prototype of miniature aerial robot, the aerial manipulator, able to accomplish operations requiring the physical interaction with the surrounding environment while remaining completely airborne. The robot arises from the combination of a vertical take-off and landing aircraft, in particular a ducted-fan configuration, and a miniature robotic arm. The physical interaction relies on impedance control considerations. The control law is able to govern all the degrees of freedom of the system both in free-flight and during the interaction with the surrounding environment. In the latter case, the stability of the contact is robustly preserved. The video shows the robotic setup and the effectiveness of the proposed control algorithm in a real-world scenario.

In this paper, a two-layer approach is presented to guarantee the stable behavior of bilateral telemanipulation systems in the presence of time-varying destabilizing factors such as hard contacts, relaxed user grasps, stiff control settings, and/or communication delays. The approach splits the control architecture into two separate layers. The hierarchical top layer is used to implement a strategy that addresses the desired transparency, and the lower layer ensures that no “virtual” energy is generated. This means that any bilateral controller can be implemented in a passive manner. Separate communication channels connect the layers at the slave and master sides so that information related to exchanged energy is completely separated from information about the desired behavior. Furthermore, the proposed implementation does not depend on any type of assumption about the time delay in the communication channel. By complete separation of the properties of passivity and transparency, each layer can accommodate any number of different implementations that allow for almost independent optimization. Experimental results are presented, which highlight the benefit of the proposed framework.

In this paper, a systematic procedure for the definition of the dynamical model in port-Hamiltonian form of mechanical systems is presented as the result of the power-conserving interconnection of a set of basic components (rigid bodies, flexible links, and kinematic pairs). Since rigid bodies and flexible links are described within the port-Hamiltonian formalism, their interconnection is possible once a proper relation between the power-conjugated port variables is deduced. These relations are the analogous of the Kirchhoff laws of circuit theory. From the analysis of a set of oriented graphs that describe the topology of the mechanism, an automatic procedure for deriving the dynamical model of a mechanical system is illustrated. The final model is a mixed port-Hamiltonian system, because of the presence of a finite-dimensional subsystem (modeling the rigid bodies) and an infinite-dimensional one (describing the flexible links). Besides facilitating the deduction of the dynamical equations, it is shown how the intrinsic modularity of this approach also simplifies the simulation phase.

In this paper, two problems related to the simulation of virtual environments for haptic systems are considered. The first problem is how to simulate, in discrete time and with low computational effort, dynamic systems in order to preserve their passivity properties. As a matter of fact, simulation of complex systems in real time may lead to undesired effects, like unstable behaviours of the haptic interface, if proper care is not given to the definition of the simulation algorithm. An algorithm is presented here able to maintain the passivity properties of the physical (simulated) system with a reduced computational complexity. The second problem discussed in this paper is the interconnection of algorithms running at different frequencies, i.e., the control algorithm of the haptic interface (running typically at high frequency) and the algorithm simulating the virtual environment (running at lower frequency). A proper software interface, able to connect these two algorithms in an energetic-consistent manner, is presented and discussed. The general framework of both these techniques is the passivity theory and the so-called port-Hamiltonian formalism.

In this paper, a simple way to model flexible robotic links is presented. This is different from classical approaches and from the Euler-Bernoulli or Timoshenko theory, in that the proposed model is able to describe large deflections in 3D space and does not rely on any finite-dimensional approximation (e.g., modal approximation). The model has been formulated within the port Hamiltonian formalism because intuitive considerations on the geometric behavior of the elastic link naturally define a Stokes-Dirac structure, the kernel of a port Hamiltonian system. Moreover, port Hamiltonian systems can be easily interconnected, thus allowing the description of complex systems as a composition of parts in an object-oriented way. By combining rigid bodies, springs, dampers, joints and, finally, flexible links, it is virtually possible to model and mathematically describe whatever complex mechanical structure formed by beams. In order to demonstrate the dynamical properties of the model and how complex mechanisms can be obtained by port interconnection, simulations of 1-DoF and 2-DoF serial manipulators and of a 2-DoF flexible closed kinematic chain are presented.

In this paper, the port Hamiltonian model of a manipulator is presented as the result of the power-conserving interconnection of a set of main components (rigid bodies, flexible links and kinematic pairs). Since rigid bodies and flexible links are described within the port Hamiltonian formalism, their interconnection is possible once a proper relation between the power conjugated port variables is deduced. These relations are the analogous of the Kirchoff laws of circuit theory. The final model is a mixed port Hamiltonian system because of the presence of a finite dimensional subsystem modelling the rigid bodies and of an infinite dimensional one describing the flexible links. The intrinsic modularity of the approach simplifies the model deduction and simulation, while the Hamiltonian description suggests the development of energy-based controllers

In this paper, the dynamics of a piezoelectric material is presented within the new framework of multi-variable distributed port Hamiltonian systems. This class of infinite dimensional system is quite general, thus allowing the description of several physical phenomena, such as heat conduction, elasticity, electromagnetism and, of course, piezoelectricity. The key point is the generalization of the notion of finite dimensional Dirac structure in order to deal with an infinite dimensional space of power variables. In this way, the dynamics of the system results from the interconnection of a proper set of elements, each of them characterized by a particular energetic behavior, while the interaction with the environment is described in terms of mechanical and electrical boundary ports.