1,720,985 research outputs found
A passivity-based adaptive admittance control strategy for physical human-robot interaction in hands-on tasks
No full textIn hands-on tasks, operator arm and robot are linked together, forming a unique dynamical system whose stability is crucial to guarantee a safe and comfortable human- robot interaction. To this extent, a passivity-based adaptation strategy is here proposed, allowing to modify the parameters of the robot control system, in accordance with changes in the operator arm impedance, so that stability is guaranteed. Experimental results, including a comparison with the methodology introduced in [1], demonstrate the importance of considering human arm impedance changes in the adaption strategy, in order to guarantee stability without causing excessive changes in admittance filter parameters
LFT-based identification of lateral vehicle dynamics
Full text linkThis paper presents a novel estimation and identification approach for lateral vehicle dynamics. The algorithm leverages on a Linear Fraction Transform (LFT) reformulation of vehicle and tyre models, allowing for a simple and computationally efficient inclusion of complex and nonlinear dynamic models, like, for example, two-wheels, four-wheels or single-track as vehicle model, and Pacejika, brush or Fiala as tyre model. As a result, this technique can be easily adopted in the development of an online identification system, able to run on a standard embedded device, implementing a flexible identification procedure that can handle different driving conditions, up to the limits of handling, different vehicle modelling approaches, and different input measurements. Experimental results demonstrate the effectiveness of the proposal, either in a persistent excitation and in a non-persistent excitation scenario
Modeling and simulation of a spherical vehicle for underwater surveillance
No full textThis paper presents the modeling and simulation of a spherical autonomous underwater vehicle. The robot was developed under the European Union H2020 innovation action UNEXMIN for the exploration of underground flooded mines, and is a small spherical robot with thrusters and an internal pendulum for pitch control.A model of the vehicle is presented, initially without the pendulum, then an extended formulation is derived accounting for a multibody dynamic description of the system. Experimental identification results for the determination of drag parameters are presented as well.A Modelica based simulator is developed for dynamic simulation of the vehicle, and is integrated with the Matlab/Simulink environment. The simulator is then validated based on preliminary experimental results
A simple and reliable technique to design kinematic-based sideslip estimators
No full textThis paper proposes two novel vehicle sideslip estimators, that aim at achieving ease of implementation and tuning, low computational cost and robustness, using only the most common automotive measurements, like vehicle position, acceleration and rotational velocity. The two estimators are only based on the unicycle kinematic model, thus they do not require any knowledge of uncertain or time-varying parameters, like vehicle parameters, or of road conditions, as it usually happens when dynamic models are adopted, and they have been derived by recasting an estimation problem into a linear control problem. Different experiments, ranging from standard driving manoeuvres to drifting driving and autonomous driving, have been performed to demonstrate the effectiveness of the proposal even in particularly critical scenarios, like driving at the limits of vehicle's handling. A comparison with a state-of-the-art sideslip estimator, using simulation and experimental data, is presented, as well
An experimentally validated LQR approach to autonomous drifting stabilization
No full textThis paper presents a drifting stabilizing controller for a rear-wheel-driven car, leveraging on front tyre steering angle and longitudinal force developed by rear tyres, the same control inputs available to a human driver. The proposed controller is based on a linear-quadratic regulator designed on a linearised single-track model of the vehicle, so that both longitudinal and lateral velocities along with yaw rate are stabilized. The controller has been experimentally validated on a scaled car. An extensive experimental campaign has been performed to demonstrate the robustness of this approach along with its shortcomings, that will be addressed in future works
A feedback linearisation algorithm for single-track models with structural stability properties
Full text linkTo support the design of trajectory tracking controllers, by way of linear tools, for vehicles characterised by nonholonomic constraints, this paper proposes a feedback linearising law for single-track dynamic models. The main characteristics of this algorithm are its simplicity, its independence from any vehicle model parameter, apart from the position of the center of mass, and its robustness. A numerical bifurcation analysis demonstrates that, for physically meaningful values of the center of mass deviation, the equilibrium is structurally asymptotically stable. Experimental results, concerning the linearising law and its application as inner loop of a trajectory tracking controller, are also presented, confirming the efficacy of the proposal
Homotopy aware kinodynamic planning using RRT-based planners
No full textThis paper introduces a method for kinodynamic planning with homotopy class constraints, and proposes a homotopy class identifier that establishes a geometric relation between a trajectory and a union of convex partitions of the 2D robot workspace. The proposed identifier is shown to be invariant with respect to the trajectories that belong to the same homotopy class, in such a way that each class has its own unique signature. Furthermore, we show that the proposed homotopy class identifier can be easily incorporated in a RRT-based planner, without changing the planning algorithm, while restricting the solution trajectories to a designated homotopy class
Object-oriented modelling of a tracked vehicle for agricultural applications
Full text linkMobile robots area pillar of technologies that support agriculture 4.0, tracked mobile robots in particular are very suitable for agricultural applications. In this paper, the development of the multibody model of a tracked vehicle for agricultural applications is described, with the main goal to create a tool for the design of motion planning and control strategies in realistic conditions, in particular taking into account the characteristics of the terrain (terramechanics). Thanks to the modular approach enabled by the object-oriented modelling language Modelica, the complete suspension system as well as the interaction of tracks and grousers with the terrain have been described with a high level of accuracy. An extensive campaign of validation experiments has been carried out, mainly focused on the steering dynamics: on a rigid surface in indoor experiments and on a deformable surface in outdoor experiments. In indoor experiments the actual position of robot center of mass and its angular orientation were measured very accurately through a VICON motion capture system and an Inertial Measurement Unit (IMU), in outdoor experiments the motion capture system was replaced by a Real-Time Kinematic (RTK) Global Positioning System (GPS). This validation campaign has demonstrated the accuracy of the proposed model in representing the vehicle dynamics, clearly stating that it can be profitably adopted as a tool to design and validate novel planning and control strategies for agricultural robots
A Leader-Follower Strategy with Distributed Consensus for the Coordinated Navigation of a Team of Quadrotors in an Environment with Obstacles
Full text linkThis paper proposes a novel strategy for the coordination of a team of quadrotor unmanned aerial vehicles (UAVs) that has to reach a target area, while moving in an environment with obstacles. The strategy consists of a coordi-nation phase, where the quadrotors assume a given formation, followed by a mission phase, where the UAV formation navigates to the target. In the latter phase, according to a leader-follower configuration, the leading agent receives from the ground station an obstacle-free trajectory to track, whereas the coordinated followers reconstruct and track their collision-free reference trajectories via a distributed consensus scheme. Finally, some simulation results are presented
Using motion primitives to enforce vehicle motion constraints in sampling-based optimal planners
No full textThis paper presents a newly conceived planning algorithm that is based on the introduction of motion primitives in RRT*. Online computational complexity of RRT* is greatly reduced by pre-computing the optimal constrained trajectories joining pairs of starting and destination configurations in a grid space, while taking into account vehicle motion constraints in the planning task. A numerical example shows the effectiveness of the algorithm
- …
