David Corinaldi

Posture Optimization of a Functionally Redundant Parallel Robot

The use of parallel-kinematics machines (PKM) for manufacturing operations is attractive because of the high accuracy they can ensure. These robots might perform a task that requires less degrees of freedom than those offered by the robot. This is the case of a robot facing a functional redundancy, which can be exploited to further increase the accuracy of the task, e.g. upon minimizing the condition number of the Jacobian matrix. A practical case study of a spherical manipulator performing a pointing task are reported, to show how posture-optimization can be used as a redundancy-resolution means for functionally redundant PKMs. The kinematics of the machine and the orientation of the pointing task is used to build, respectively, the objective function and the constraint equations. Sequential Quadratic Programming is conducted to solve the nonlinear constrained optimization problem and to find the end-effector pose corresponding to the robot posture of minimum condition number for every direction of a given pointing path. Lastly, the constrained problem is rewritten as one of unconstrained optimization of one objective function in one design variable.

Sensitivity analysis of a mini pointing device

The paper presents a preliminary study needed to carry out a kinematic calibration procedure for a mini pointing device. The latter inherits its kinematics from a conventional five-bar linkage. A sensitivity analysis of all the geometric parameters involved in the kinematic model of the device is performed within the device workspace. A model is proposed by assuming a non overconstrained kinematics for the machine. Such assumption allows to consider a coupled rotational and translational motion of its moving platform, that is usually designed to have a fixed center of rotation. Results show how the model can be simplified without a significant reduction of its position accuracy, at least in a significant region of the manipulator workspace.

Experimental analysis of a fractional-order control applied to a second order linear system

This paper presents the application of a PDD1/2 fractional order controller to a purely inertial second order system, with the aim of deepening the study of the half-derivative contribution on control performances, in particular on settling time and settling energy of a step response. An approximation of the half-derivative term is proposed for a discrete time version of the controller. The choice is justified by the results obtained by an analysis of order and sampling time influence on accuracy and computation time of a discrete-time fractional derivative. Afterwards, simulation results, all obtained in dimensionless form, are extended to the physical case by means of experimental tests on a rotary axis. In the end it is shown that the half-derivative approximation still allows to improve the system dynamics if compared with a conventional PD controller.

Sensitivity Analysis and Model Validation of a 2-DoF Mini Spherical Robot

This paper is focused on the development and validation of an error kinematic model of a mini spherical robot, aimed at its kinematic calibration. The robot is actually a spatial five-bar linkage, provided with two rotational degrees of freedom. A non-overconstrained kinematics is assumed for the robot in order to provide a simple mathematical model and allow a sensitivity analysis of all the involved geometric parameters. A simplified version of the model is proposed. It differs only for the degree of approximation. A comparison between full and reduced models is presented by means of numerical simulations, analyzing their behavior in a significant region of the robot workspace. In order to validate both of them a robot calibration is carried out on a physical prototype of the robot using a vision system, namely a fixed camera in a eye-to-hand configuration. An iterative algorithm aimed at the experimental identification of the geometric data of the robot is used. Some experimental results show the effectiveness of the study.

Rotational Mobility Analysis of the 3-RFR Class of Spherical Parallel Robots

Spherical parallel manipulators (SPMs) are used to orient a tool in the space with three degrees of freedom exploiting the strengths of a multi-limb architecture. On the other hand, the performance of parallel kinematics machines (PKMs) is often affected by the occurrence of different kinds of singular configurations. The paper aims at characterizing a class of SPMs for which all singularities come to coincide and a single expression is able to describe all the singular configurations of the machines. The study is focused on a class of SPMs with 3-RFR topology (Revolute-Planar-Revolute pairs for each of the three limbs) addressing the mobility and singularity analysis by means of polynomial decomposition and screw theory. The neatness of the equations that are worked out, expressed in a robust formulation based on rotation invariants, allows a straightforward planning of singularity free tasks and simplifies the synthesis of dexterous machines.

A Novel Reconfigurable 3-URU Parallel Platform

A novel parallel robot stemming from the 3-SRU (spherical-revolute-universal) under - actuated joints topology is presented in this paper. The conceptual design here proposed takes advantage of a reconfigurable universal joint obtained by locking, one at a time, different rotations of a spherical pair by means of an automatized system. Such local reconfiguration causes a slight modification of the robot legs mobility which is enough to provide the end-effector with different kinds of motion. The first part of the paper is dedicated at formally demonstrating the motion capabilities offered by the 3-URU kinematics; in the second part, a mechanical solution for the realization of the reconfigurable joint is shown.

Synthesis of a Spatial 3-dof Deployable Mechanism to Grasp Stacked Non-Rigid Materials

The paper deals with the synthesis of a novel deployable mechanism with three decoupled degrees of freedom for the handling of large plies of non-rigid material stacked on a beam. The gripper structure is made up of a repeated deployable unit. This basic mechanism is an assembly of Sarrus and scissor linkage: the first to move the unit in three independent directions and the other to transmit the motion to adjacent elements. The particular shape of the beam requires a symmetrical transmission of motion solved by adding linkages of the same type. The final assembly grants the uncoupling of the three degrees of freedom, which allows an anisotropic scaling of the whole device on a flat piecewise surface. The kinematic analysis highlights the feasibility and effectiveness of the described scale transformations of the structure and simulations are performed to verify the deployability.

A Gripper for Handling Large Leather Plies Stacked on Beams

The paper presents the preliminary design of a novel gripper able to grasp large non-rigid materials that has been conceived to face the challenge of automatic handling tasks in the leather industry. The design has been driven by the requirements to limit production costs and the complexity of the grasping device. A statistical analysis of the different templates sizes has allowed to identify a fixed configuration of the gripping points able to properly pick all the sheets within a great confidence interval. According to the varying shape of the leather templates themselves, that is due to their stacking in plies on the beam, the trajectory of the gripping points has been studied and arranged. Due to the irregular shape of the large sheets that are handled, the edges of the non-rigid materials out of the gripping area might flutter during the transferring phase: a four-bar linkage has been specifically designed, so that the motion of its end-effector prevents unwanted leather creases.Copyright