Scribd
Upload
8 views4 pages

App 8

Uploaded by

alenelijalem12
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
8 views4 pages

App 8

Uploaded by

alenelijalem12
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 4

Appendix 8

Computation of the inertia matrix of tree


structured robots

In this appendix, we develop a method to compute efficiently the inertia matrix of tree
structured robots. Note that a serial robot is a special case of the tree structured robot. This
method is based on the utilization of a simplified special case of Newton-Euler algorithm and
on the concept of composite links [Khalil 90b].

A8.1. Inertial parameters of a composite link

The composite link j"** is composed of link j and of the links supported by link j (Figure
A8.1). The inertial parameters of the composite link j"^ can be calculated in terms of the
standard parameters (or base parameters) of its links using the following recursive algorithm:

i) initialization. For J =/,...,«.•

i j * = ijj, JMSJ = iMSj, M^ = Mj

We recall that a(j) indicates the link that is antecedent to link j ;

ii)forj = n,..., 2 and a(j) ^ 0:

+ »(J)Pj«(J)pjMj* (A8.1al
432 Modeling, identification and control of robots

with:
- »0)MSJ^ = »(J)AjiMsJ';
- v: (3x3) skew-symmetric matrix of the components of the vector v;
aO)Aj «(J)Pj
- aO)Tj =
L 03,1 '
- JJj: inertia tensor of the composite link j * referred to frame RJ;
- JMS::firstmoments of the composite link j"** referred to frame RJ;
- M:: mass of the composite link j"*".

We note that equations [A8.1] are equivalent to the following:


IA8.2]

where
NOTE.-^^^Xj
TheandJKj are defined
relationship in Chapters
between 9 and
the concept 10 and Appendix
of composite 7.
link parameters and base inertia!
parameters is considered in [Khalil 90a].

Figure A8.1. Composite linkj^


Inertia matrix of tree structured robots 433

A8.2. Computation of the inertia matrix

We have seen in § 9.7.1 that the j^** column of the inertia matrix A can be computed by
the Newton-Euler inverse dynamic algorithm by setting:

ij = Uj, q = 0, g = 0, Fc = 0 (fci = 0, mei = 0 for i = 1 n)

where Uj is an (nxl) vector with a 1 in the j^** row and zeros elsewhere.

Under these conditions, the forward recursive equations of the Newton-Euler inverse
dynamic (Chapter 9) are only applied to link j"*":

*^CDk =: 0, *^d) ic = 0, ^Vk = 0, ^Fk = 0, ^Mk = 0 for k < j [A8.3]


i(Dj = 0 (A8.4]
i(bj = SjJaj [A8.5]

iVj = Ojiaj [A8.6]

JFj = M J iVj -f Jd)j X JMSJ [A8.7]

JMj = i j j id)j + iMsJ X iVj [ A8.8]

We deduce that:
- if joint j is prismatic (icbj« 0, JMj« 0 and J Vj = [ 0 0 I ] \ then:

JFj = [0 0 Mjf [A8.91


JMj = (MYJ" -MX]^ o f [A8.10]

- if joint j is revolute (iVj = 0 and Jcbj = [ 0 0 1 ] \ then:

iFj = hMY^ MXj"" of [A8.I1)


iMj = [XZj YZ| izjf [A8.12]

The recursive backward computation starts by link j and ends with link s» where a(s) = 0.
The algorithm is given by the following equations:
- if joint j is prismatic, then:

jfj = JFj = [0 0 M]^f [A8.13]


434 Modelings identification and control of robots

iiT
Jnij = JMj = [MYj -MXj 01' [A8.14]
,+ +Iaj
Ajj = Mj [A8.15]

- if joint j is rcvoIutc» then:

ifj = JFj = [-MY/ Mx/ of [A8.16)

Jitij = JMj = [XZ/ vz; Tzjf [A8.17J

Ajj = Z z | + Iaj (A8.18]

Then, the following equations are computed for k=:j. a(j), a(a(j),..., s. where a(s)=0:

a(k)f^(,,^ = a(k)A^kf^ [A8.19]

ra(k) = Aa(k),j = (C^a(k)^a(k) + 5a(k)ina(k))^*^^^«a(k) [A8.21]

NOTES-
- the element Aj j of the inertia matrix is set to zero if link i does not belong to the path
between the base and link j ;
- this algorithm provides the elements of the lower part of the inertia matrix. The other
elements are deduced using the fact that the inertia matrix A is symmetric.

You might also like