Recapitulation: Forces and Torques for static manipulators

Similar documents
Lecture 7. Matthew T. Mason. Mechanics of Manipulation. Lecture 7. Representing Rotation. Kinematic representation: goals, overview

Lecture L3 - Vectors, Matrices and Coordinate Transformations

ME 115(b): Solution to Homework #1

TWO-DIMENSIONAL TRANSFORMATION

Figure 1.1 Vector A and Vector F

Physics 235 Chapter 1. Chapter 1 Matrices, Vectors, and Vector Calculus

December 4, 2013 MATH 171 BASIC LINEAR ALGEBRA B. KITCHENS

Rotation Matrices and Homogeneous Transformations

Mathematics Course 111: Algebra I Part IV: Vector Spaces

Design-Simulation-Optimization Package for a Generic 6-DOF Manipulator with a Spherical Wrist

Mechanics lecture 7 Moment of a force, torque, equilibrium of a body

Question 2: How do you solve a matrix equation using the matrix inverse?

Operational Space Control for A Scara Robot

Unified Lecture # 4 Vectors

Vector Algebra II: Scalar and Vector Products

NEW YORK STATE TEACHER CERTIFICATION EXAMINATIONS

Solving simultaneous equations using the inverse matrix

The Matrix Elements of a 3 3 Orthogonal Matrix Revisited

2 Session Two - Complex Numbers and Vectors

Essential Mathematics for Computer Graphics fast

Chapter 17. Orthogonal Matrices and Symmetries of Space

Introduction to Engineering Analysis - ENGR1100 Course Description and Syllabus Monday / Thursday Sections. Fall '15.

13 MATH FACTS a = The elements of a vector have a graphical interpretation, which is particularly easy to see in two or three dimensions.

INSTRUCTOR WORKBOOK Quanser Robotics Package for Education for MATLAB /Simulink Users

Matrix Differentiation

by the matrix A results in a vector which is a reflection of the given

F Matrix Calculus F 1

LS.6 Solution Matrices

8.2. Solution by Inverse Matrix Method. Introduction. Prerequisites. Learning Outcomes

Rotation about an arbitrary axis and reflection through an arbitrary plane

Let s first see how precession works in quantitative detail. The system is illustrated below: ...

Linear Algebra Notes for Marsden and Tromba Vector Calculus

Vectors VECTOR PRODUCT. Graham S McDonald. A Tutorial Module for learning about the vector product of two vectors. Table of contents Begin Tutorial

MATRIX ALGEBRA AND SYSTEMS OF EQUATIONS

A linear combination is a sum of scalars times quantities. Such expressions arise quite frequently and have the form

Lecture 3: Coordinate Systems and Transformations

discuss how to describe points, lines and planes in 3 space.

Torgerson s Classical MDS derivation: 1: Determining Coordinates from Euclidean Distances

Continued Fractions and the Euclidean Algorithm

The elements used in commercial codes can be classified in two basic categories:

2. Spin Chemistry and the Vector Model

3D Tranformations. CS 4620 Lecture 6. Cornell CS4620 Fall 2013 Lecture Steve Marschner (with previous instructors James/Bala)

Notes on Orthogonal and Symmetric Matrices MENU, Winter 2013

MECHANICAL PRINCIPLES HNC/D MOMENTS OF AREA. Define and calculate 1st. moments of areas. Define and calculate 2nd moments of areas.

Design of a six Degree-of-Freedom Articulated Robotic Arm for Manufacturing Electrochromic Nanofilms

2x + y = 3. Since the second equation is precisely the same as the first equation, it is enough to find x and y satisfying the system

Dynamics. Basilio Bona. DAUIN-Politecnico di Torino. Basilio Bona (DAUIN-Politecnico di Torino) Dynamics / 30

DERIVATIVES AS MATRICES; CHAIN RULE

Content. Chapter 4 Functions Basic concepts on real functions 62. Credits 11

Solving Systems of Linear Equations

Given a point cloud, polygon, or sampled parametric curve, we can use transformations for several purposes:

ASEN Structures. MDOF Dynamic Systems. ASEN 3112 Lecture 1 Slide 1

Introduction to Matrix Algebra

MATRIX ALGEBRA AND SYSTEMS OF EQUATIONS. + + x 2. x n. a 11 a 12 a 1n b 1 a 21 a 22 a 2n b 2 a 31 a 32 a 3n b 3. a m1 a m2 a mn b m

Introduction to Matrices for Engineers

9.4. The Scalar Product. Introduction. Prerequisites. Learning Style. Learning Outcomes

A vector is a directed line segment used to represent a vector quantity.

Unit 18 Determinants

Simulation of Trajectories and Comparison of Joint Variables for Robotic Manipulator Using Multibody Dynamics (MBD)

Basic CAD/CAM. CHAPTER 5: Geometric Transformation

Geometry of Vectors. 1 Cartesian Coordinates. Carlo Tomasi

Solving Simultaneous Equations and Matrices

Geometric Transformations

Lecture L22-2D Rigid Body Dynamics: Work and Energy

Force/position control of a robotic system for transcranial magnetic stimulation

Typical Linear Equation Set and Corresponding Matrices

Current Standard: Mathematical Concepts and Applications Shape, Space, and Measurement- Primary

521493S Computer Graphics. Exercise 2 & course schedule change

The Characteristic Polynomial

How To Understand The Dynamics Of A Multibody System

Constraint satisfaction and global optimization in robotics

Matrices 2. Solving Square Systems of Linear Equations; Inverse Matrices

Lectures notes on orthogonal matrices (with exercises) Linear Algebra II - Spring 2004 by D. Klain

Synthesis of Constrained nr Planar Robots to Reach Five Task Positions

Systems of Linear Equations

[1] Diagonal factorization

4.5 Linear Dependence and Linear Independence

Lecture 2 Matrix Operations

ANALYSIS OF STRUCTURAL MEMBER SYSTEMS JEROME J. CONNOR NEW YORK : ':,:':,;:::::,,:

Section V.3: Dot Product

Vector has a magnitude and a direction. Scalar has a magnitude

The Basics of FEA Procedure

National 5 Mathematics Course Assessment Specification (C747 75)

The Point-Slope Form

Equations, Inequalities & Partial Fractions

Véronique PERDEREAU ISIR UPMC 6 mars 2013

3.2 Matrix Multiplication

DRAFT. Further mathematics. GCE AS and A level subject content

LINEAR ALGEBRA W W L CHEN

Algebra 2 Chapter 1 Vocabulary. identity - A statement that equates two equivalent expressions.

Mathematics Pre-Test Sample Questions A. { 11, 7} B. { 7,0,7} C. { 7, 7} D. { 11, 11}

CITY UNIVERSITY LONDON. BEng Degree in Computer Systems Engineering Part II BSc Degree in Computer Systems Engineering Part III PART 2 EXAMINATION

EDEXCEL NATIONAL CERTIFICATE/DIPLOMA MECHANICAL PRINCIPLES AND APPLICATIONS NQF LEVEL 3 OUTCOME 1 - LOADING SYSTEMS

3 Orthogonal Vectors and Matrices

Linear Algebra: Vectors

MATHEMATICS FOR ENGINEERING BASIC ALGEBRA

Similarity and Diagonalization. Similar Matrices

MATHEMATICS FOR ENGINEERS BASIC MATRIX THEORY TUTORIAL 2

Understanding and Applying Kalman Filtering

1 Introduction to Matrices

Transcription:

Recapitulation: Forces and Torques for static manipulators For propagation of forces and torques in a non-moving manipulator, the following equations hold: i f i = i i+r i+ f i+ i n i = i i+r i+ n i+ + i P i+ i f i The force that affects a link is denoted by f (a three-dimensional vector), and the torque on that link is denoted by n (also three-dimensional). Note that some parts of the forces and torques apply directly to the corresponding joint, and some parts are absorbed by the mechanics of the robot. The relation between these quantities is: τ i = i n T i i Z i = i n T i τ i = i f T i i Z i = i f T i The first equation is used for rotational joints, the second equation for prismatic joints. The quantities τ i thus specify the amount of torque resp. force that is affecting the joint, and thus the amount of torque resp. force that the robot should counteract in order to remain static. The joint torques/forces τ i are -dimensional quantities. The Jacobian plays a very important role here as well: It relates joint torques/forces τ i to endeffector forces and torques f, n: τ τ. = τ = A J T A F = A J T τ n ( A ) f A n Where F is a 6-dimensional Vector of force and torque containing 3D force and 3D torque vectors stacked above each other. It is important that the force-torque vector has the same frame of reference as the Jacobian. Rotational velocities and the Jacobian In the previous problem sheet, entries of the Jacobian concerning rotations have been computed based on simplified position descriptions. We have not explained yet how the Jacobian can be computed efficiently in the general case where arbitrary rotations are possible. Let p : R n R 3 be a function that computes the coordinates of the origin of the end effector with respect to system {}, then the full Jacobian looks like this: p p p x x... x n p p p J = x x... x n p 3 p 3 p x x... 3 x n Ẑ Ẑ... Ẑ n We are familiar with the first three rows of this Jacobian, but the final row is something new. First of all, we need to explain the meaning of j Ẑ i. This denotes the z-axis of system i, expressed relative

Figure : 3R-Robot to frame {j}. Thus, it is a three-dimensional unit vector, and we have i Ẑ i = (,, ) T. So we see that the last row in the above matrix really stands for three rows. An example: For the well-known planar 3R-manipulator from the previous exercises, the last three rows of the complete Jacobian would look like this: J = ( ) Ẑ Ẑ Ẑ 3 = This is obvious, since all joint axes are parallel and represent rotations around Ẑ, and it also corresponds to the observation that joint rates simply add up. Now what if the third axis were rotated about 9 instead of being parallel to the second axis? The corresponding robot is shown in Figure. According to above rules, the lower part of the Jacobian would look like this: J = ( ) s Ẑ Ẑ Ẑ 3 = c This observation often simplifies the computation of the Jacobian: All you have to do is to determine the direction of the z-axes of the coordinate systems. Note however, that a prismatic joint will never generate a rotational velocity, thus the entry in the column corresponding to the rotational velocity of a prismatic joint is set to (,, ) T instead of Ẑ i. If determining the directions of the joint axes becomes too complicated, you can still apply the formulas for computing rotational velocities i ω i and derive the Jacobian from that. If you do that, always keep in mind that you need to be in the right frame of reference, i.e., if the position description is in frame, you need to transform n ω n to frame as well if you want to compute J. The mathematical justification for above rule is as follows: For the computation of ω, ω,... there is a recursive formula i+ ω i+ = i+ i R iω i + Θ i+ i+ Ẑ i+.

If you apply this formula generally to n ω n, a certain pattern becomes visible: n ω n = n n R n ω n + Θ n nẑ n = n n R ( n n R n ω n + Θ n n Ẑ n ) + Θ n nẑ n = n n R ( n n R ( n n 3R n 3 ω n 3 + Θ n n Ẑ n ) + Θ n n Ẑ n ) + Θ n nẑ n = n R Θ + n R Θ + n 3R Θ 3 +... + n nr Θ n We are only interested in ω n, thus we further compute: ω n = nr n ω n = nr n R Θ + nr n R Θ + nr n 3R Θ 3 +... + nr n nr Θ n = R Θ + R Θ + 3R Θ 3 +... + nr Θ n = Θ Ẑ + Θ Ẑ + Θ 3 Ẑ 3 +... + Θ n Ẑ n Collecting the factors of Θ, Θ, Θ 3,..., we see that the entries correspond to Ẑ, Ẑ and so on, and we see that the scheme for computation of the Jacobian is correct! Solution a) The Jacobian relates joint torques to external forces as follows: τ = 4 J T 4 f Here, we have 4 joint parameters, and the formula amounts to τ τ τ 3 = (4 J T ) 4 f = ( J T ) f. τ 4 To determine the Jacobian (according to the recipe in the recap.), we need to compute P 4ORG. The transformation matrices are: cos Θ sin Θ cos Θ sin Θ T = sin Θ cos Θ, T = sin Θ cos Θ cos Θ 3 sin Θ 3 cos Θ 4 sin Θ 4 sin Θ 3 cos Θ 3 3T = sin Θ 3 cos Θ 3, 3 4T = sin Θ 4 cos Θ 4

Based on that we can compute T T 3T 3 4T (,,, ) T step by step, which results in: sin (Θ + Θ ) sin Θ 3 + cos (Θ + Θ ) cos Θ 3 + sin (Θ + Θ ) + cos Θ p(θ) = cos (Θ + Θ ) sin Θ 3 + sin (Θ + Θ ) cos Θ 3 cos (Θ + Θ ) + sin Θ sin Θ 3 + The entries of the Jacobian, evaluated for Θ = (, 9, 9 ), are then: p Θ = s c 3 c s 3 + c s = + = p Θ = s c 3 c s 3 + c = + = p Θ 3 = c s 3 s c 3 = = p Θ = c c 3 s s 3 + s + c = ( ) + + = 3 p Θ = c c 3 s s 3 + s = + + = p Θ 3 = s s 3 + c c 3 = + = p 3 Θ = p 3 Θ = p 3 Θ 3 = c 3 = Thus we have computed the first three rows of the Jacobian. The lower 3 rows can be determined according to above considerations as Ẑ, Ẑ, Ẑ 3, Ẑ 4 : (,, ) T, (,, ) T, (,, ) T, (,, ) T. All in all, the transposed Jacobian looks like this: 3 J T = Since the force-torque-vector is specified relative to system 4, we need to transform it to system. This is done by multiplying with the corresponding rotation matrix: ( 4 ) ( R 4 ) 6 f = 3 3 4 R f = 3 3 3 3 4 R 3 3 4 R 7 8 The rotation matrix 4R in the current configuration Θ = (, 9, 9, ) T is R R 3R 3 4R = I I =.

Since we already have determined the direction of the z-axis of system {4}, we can also derive the rotation matrix 4R by figuring out ˆX4, Ŷ 4 in addition to Ẑ 4 : ˆX4 =, Ŷ 4 =, Ẑ 4 = The vectors are the columns of the rotation matrix. All in all, we obtain: f = Finally, we can apply the formula for computing joint moments: 6 5 3 τ = 6 5 8.77 =.77 6.485 8 b) Now we want to compute the forces and torques at the tip of a screwdriver that is attached to the robot. If we denote the system of the screw driver s tip with {5}, we have: 4 f 4 = 4 5R 5 f 5 5 f 5 = 4 f 4 7 6 4 n 4 = 4 5R 5 n 5 + 4 5P 4 f 4 5 n 5 = 4 n 4 4 5P 4 f 4 = 6 = 8 9 8 The force that the robot causes in direction of the screw driver (along 4 Ẑ 4 = 5 Ẑ 5 ) is equal to the third component of 5 f 5, because: 4 Ẑ4 T 6 = 6 = Thus, the robot does not apply any force at all in that direction. Furthermore, along the same axis, a torque of 6 4 Ẑ4 T = 8 8 is caused by the robot. T

Figure : Choice of coordinate systems Solution Let s start by computing the Jacobian based on velocities. The formulas needed for the computation have been summarized in solution. For application of these formulas, we need to determine the rotation matrices between the systems. To determine these rotation matrices, we first need to apply the DH convention to determine coordinate systems and DH parameters. Figure shows the coordinate systems. The robot is shown in configuration Θ = Θ = Θ 3 =, and the DH parameters are now: i a i α i d i Θ i Θ l 9 Θ 3 l Θ 3 (4) l 3 With these specifications, the rotation matrices are determined as: c s c s l T = s c T = s c c 3 s 3 l l 3 3T = s 3 c 3 3 4T = Note the additional transformation 3 4T that is only a translation along 4 x to the endeffector system. The rotation matrices i+ i R are equal to the transpose of the 3 3 rotation part of the transformation

matrices. The velocities compute as follows: ω = R + Θ = Θ v = R ( + P ) = s Θ ω = R + = c Θ Θ Θ Θ v = R + l = R Θ l Θ = l Θ s Θ s Θ c 3 + c Θ s 3 3 ω 3 = 3 R c Θ + = s Θ s 3 + c Θ c 3 Θ Θ 3 Θ + Θ 3 s Θ s 3 Θ l 3 v 3 = 3 R Θ + c Θ l = 3 R Θ l = l Θ l c Θ Θ c 3 Θ l l l c Θ Θ l 4 ω 4 = 3 ω 3 4 v 4 = I 3 v 3 + 3 ω 3 = 3 l = 3 v 3 + ( Θ + Θ 3 )l 3 l 3 (c Θ c 3 s Θ s 3 ) s 3 Θ l Θ (l s 3 ) = Θ (l c 3 + l 3 ) + l 3 Θ3 Θ (l 3 (s s 3 c c 3 ) l c l ) c 3 Θ l + Θ l 3 + Θ 3 l 3 l 3 s Θ s 3 l 3 c Θ c 3 l c Θ Θ l The final values 4 v 4 and 4 ω 4 can be simplified using the trigonometric identities (s s 3 c c 3 = c 3... ), and we can derive the complete Jacobian: l s 3 l c 3 + l 3 l 3 4 J = l 3 c 3 l c l s 3 c 3 Thus, we have determined the Jacobian based solely on velocities. For the following considerations, the computations will not be as exhaustive, we will more focus on the basic ideas. For the force/torque-relations we have: τ = A J T 4 F

We denote the components of 4 F as follows: Note that: F F F = N N N 3 τ i = i n i (,, ) T This is true because our robot has only rotational joints. Applying the formulas for computing forces and torques yields: 4 f 4 = 3 f 3 = F F F F 4 n 4 = 3 n 3 = F cos Θ 3 F sin Θ 3 f = F sin Θ 3 + F cos Θ 3 n = N N N 3 N N l 3 N 3 + F l 3 N cos Θ 3 (N l 3 ) sin Θ 3 N sin Θ 3 + (N l 3 ) cos Θ 3 l l (F sin Θ 3 + F cos Θ 3 ) + N 3 + F l 3 F cos (Θ 3 + Θ ) F sin (Θ 3 + Θ ) f = F sin (Θ 3 + Θ ) + F cos (Θ 3 + Θ ) ( l 3 N ) sin (Θ 3 + Θ ) + N cos (Θ 3 + Θ ) + l sin Θ n = F l sin (Θ 3 + Θ ) l F cos (Θ 3 + Θ ) F l sin Θ 3 F l cos Θ 3 N 3 F l 3 N sin (Θ 3 + Θ ) + (N l 3 ) cos (Θ 3 + Θ ) l cos Θ l We are interested in the values of τ, τ, τ 3 which are going to allow us to deduce the entries of the Jacobian. As stated above, they are the third components of i n i. Note that this means that we need only compute the third component of n, which simplifies the computation. All in all, we have now computed the following values for τ: τ = N sin (Θ 3 + Θ ) + (N l 3 ) cos (Θ 3 + Θ ) l cos Θ l τ = l (F sin Θ 3 + F cos Θ 3 ) + N 3 + F l 3 τ 3 = N 3 + F l 3 Now we collect the coefficients of F, F,, N, N, N 3, which allows us to determine the shape of the transposed Jacobian: 4 J T F F N sin (Θ 3 + Θ ) + ( l 3 cos (Θ 3 + Θ ) l cos Θ l ) + N cos (Θ 3 + Θ ) N = F l sin Θ 3 + F (l cos Θ 3 + l 3 ) + N 3 N N 3 + F l 3 N 3

l 3 c 3 l c l s 3 c 3 4 J T = l s 3 l c 3 + l 3 l 3 The geometric method for computing the Jacobian would be the differentiation of the cartesian coordinates with respect to Θ, Θ, Θ 3 for the upper 3 rows and computing the lower 3 rows as Z, Z, Z 3, as shown in Solution. We can thus determine J: p p p Θ Θ Θ 3 p p p J = Θ Θ Θ 3 p 3 p 3 p 3 Θ Θ Θ 3 Z Z Z 3 The position of the origin of system 4 is computed as: The vectors Z, Z, Z 3 are: T T 3T 3 4T(,,, ) T Z = (,, ) T Z = R(,, ) T = ( s, c, ) Z 3 = Z = ( s, c, ) Then we can use the following relation: J = ( 4 R 3 3 3 3 4 R ) ( 4 ) 4 J 4 R J = 3 3 4 J 3 3 R This relation has been established by multiplying both sides with the inverse of the compositional rotation matrices. The computation of the explicit value is pure computational work and will yield the same result as the other calculations before.