Inertial and unstability

Hi everyone,

I build a robotic arm (made of cynliders) without setting the inertials first. Everything was working well using a PID to control it. Then I set up the inertials using the list on wikipedia:

https://en.wikipedia.org/wiki/List_of_moments_of_inertia

It appears that the values I get are very small and the model became unstable (all links are exploding in every directions). The different pieces of the arm are mostly like this.

radius = 0.025
length = 0.5
mass = 0.1

So their inertial are:

ixx = (mass/12)(3radius^2+length^2)
iyy = (mass/12)(3radius^2+length^2)
izz = mass*radius^2/2

I don't get why it is not stable. So here are my questions:

1) Do you have any idea why it is not stable ?
2) Do you know what happens if we set the inertials to zero ?
3) Do you know what happens when we don't set the inertials ?

Cheers

Asked by debz on 2015-10-02 11:28:38 UTC

Comments

Answers

If you are using ODE, can you try increasing the solver iterations?

http://gazebosim.org/tutorials?tut=modifying_world&cat=build_world#PhysicsProperties

Asked by scpeters on 2015-10-02 16:14:36 UTC

Comments

Yes thx I tried but nothing changed ! I multiplied it by 10 (50->500 iterations). By the way, do you know what are the default parameters of the inertials ?

Asked by debz on 2015-10-08 13:57:30 UTC

Given this topic, thevalues I get might be too small and interpreted as zeros... Anyone has an idea how to deal with that.

One says "Unfortunately, this usually gives a matrix that is rounded to all zeroes, so you might have to scale up your model, compute the inertia tensor, and divide the result by an appropriate factor.". But if you scale up your inertials parameters, then it will affect the behaviour of PIDs right ?

Cheers

Asked by debz on 2015-10-08 14:44:33 UTC