Human & Robot Interaction Lab. (TaarLab)

Login
A+ A A-

Although classical and preliminary kinematic investigations of robotic mechanical systems are the principal goal Human and Robot Interaction laboratory, resorting to algebraic geometry is essential because problems eventually became elusive to classical approaches. The use of algebraic geometry leads to some ground-breaking results which could be astonishing. The alternation of perspectives is prominently present in the different projects.

Thus, as a global objective, Human and Robot Interaction Laboratory aims at reconciling two nearly disconnected communities, namely the geometricians and mechanical engineers. With that in mind,all the theoretical objectives about the kinetostatic analysis of robotic mechanical systems are treated by veering a little from our engineering grounding and using algebraic geometry concepts, such as seven-dimensional kinematic space (the so-called Study's parameters), Grassmann-Cayley Algebra and Screw Theory. To achieve the aforementioned global objective, the obtained results are also re-explored using engineering vision to fill some gaps that geometrical treatments fails to provide in a design context.

Singularity Analysis of Parallel Manipulators using Grassmann-Caylay Algebra

People Involved
   Arvin Rasoulzadeh, Behzad Danaei

Singularity Analysis has always been an important topic to robotics researchers and is an active research field since part of controlling a robot is to avoid the singular points. Here at Taarlab we propose the new methods of dealing with this area of research by using the modern technics borrowed from Geometrical Algebra. The efficiency of these methods are to their simplicity and being coordinate-free.

ag1 

Analysis of the forward kinematic of the 4-DOF spatial parallel robots with identical limb structures and specified motion pattern of 3T1R

People Involved
   Payam Varshovi, Davood Naderi

A parallel robot is a closed-loop kinematic chain mechanism whose moving platform is
linked to the base by at least two independent kinematic chains. Due to the resulting system of nonlinear equations, the Forward Kinematic Problem (FKP) of parallel robots is complicated and cannot be easily analyzed. This thesis investigates the FKP of 4-DOF parallel robots with identical kinematic chain structures performing three translations and one rotation motion pattern, referred to as Schönflies motion. These robots have a short history and very few studies have been conducted on their kinematic characteristics. Based on the complexity of the FKP of 4-DOF parallel robots, the proposed algorithms are implemented on 3-DOF planar parallel robots, which are simpler structures, and then they are applied on 4-DOF parallel robots. First, the FKP of the parallel robots are studied in three-dimensional Euclidean space. In this approach, the joint position, kinematic constraints and resultant method are used for analysis of the FKP of parallel robots. Moreover, for the sake of comparison, the system of equations corresponding to the forward kinematic problem is solved upon resorting to Bertini software and inverse kinematic problem. Then, the FKP of parallel robots are investigated in seven-dimensional kinematic space. In this method, the FKP is analyzed by means of Denavit-Hartenberg parameters, Study’s kinematic mapping, LIA algorithm and Gröbner basis. The solutions of the FKP in sevendimensional kinematic space are mapped to the three-dimensional Euclidean space and compared to the obtained results from resultant method. More, a new algorithm is presented in sevendimensional kinematic space by means of Euler’s parameters, LIA method and interval analysis, so that this algorithm analyzes the FKP of parallel robots considering to the motion range of passive joints. The results obtained from new algorithm are compared with the results of the presented method in seven-dimensional kinematic space and fully confirmed the validity.

ag2 

TaarLab 2014, All Rights Reserved.
Designed, Developed and Powered by: Payam Ghasemi & Nima Karbasi

Login