Input-according-to-MATLAB-inverse-function-in-Robot-Manipulator

Authors: Seonghyeon Jo(cpsc.seonghyeon@gmail.com)

Date: Feb, 07, 2022

Comparison of input from robot manipulator due to computational error for MATLAB inverse function. [0] : identity matrix

[1] : inverse matrix

(https://kr.mathworks.com/help/matlab/ref/inv.html)

[2] : Moore-Penrose pseudoinverse function

(https://kr.mathworks.com/help/matlab/ref/pinv.html)

[3] : mldivide, mrdivide function

(https://kr.mathworks.com/help/matlab/ref/mldivide.html)

(https://kr.mathworks.com/help/matlab/ref/mrdivide.html)

[4] : round, mldivide, mrdivide function

result table

1. mean(u_r - u)

Method	0	1	2	3	4
joint 1	1e-00*0.0	1e-14*0.5	1e-13*0.4	1e-14*0.1	1e-00*0.0
joint 2	1e-00*0.0	1e-15*0.8	1e-14*0.6	1e-15*0.2	1e-00*0.0

2. mean(abs(I-B))

Method	0	1	2	3	4
(1,1)	1e-00*0.0	1e-16*0.6	1e-14*0.1	1e-16*0.5	1e-00*0.0
(1,2)	1e-00*0.0	1e-16*0.1	1e-15*0.2	1e-17*0.8	1e-00*0.0
(1,3)	1e-00*0.0	1e-15*0.5	1e-14*0.4	1e-15*0.1	1e-00*0.0
(1,4)	1e-00*0.0	1e-16*0.6	1e-15*0.6	1e-16*0.1	1e-00*0.0
mean	1e-00*0.0	1e-15*0.2	1e-14*0.2	1e-16*0.5	1e-00*0.0

result figures

0. identity matrix

1. inverse function

2. Moore-Penrose pseudoinverse function

3. mldivide, mrdivide function

4. round, mldivide, mrdivide function