algebraic loop and Simscape ?

Question

aft 2011-9-26

1
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/16802-algebraic-loop-and-simscape

Hi,

I am having a bad time dealing with algebraic loops in my inverted pendulum model. Leaving it to simulink to solve using its internal solvers takes a lot of time while adding an integer delay modifies my results in an undesirable way. On the following link of Mathworks,

http://www.mathworks.fr/help/toolbox/simulink/ug/f7-8243.html#bsxw0e2-1,

on the work-flow chart they say in the last box that converting the model "Simscape" is also a solution. I am wondering HOW ? it looks to me all the same thing ?

anybody has an idea ?

thanx in advance

khawaja

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Guy Rouleau 2011-9-26

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/16802-algebraic-loop-and-simscape#answer_22673

2D and 3D physical systems like a pendulum on cart are most easily represented by a Differential Algebraic Equations (DAE), and not only a set of Ordinary Differential Equations (ODE). In Simulink, a DAE causes an algebraic loop.

The Simscape (and SimMechanics) solver is designed to solve DAE better than the Simulink algebraic loop solver. In your case, for 2D or 3D, the solution would be SimMechanics:

http://www.mathworks.com/products/simmechanics/

To understand the differences between Simulink and SimMechanics, I recommend starting with this post Seth and I made some time ago:

http://blogs.mathworks.com/seth/2009/02/26/modeling-mechanical-systems-the-double-pendulum/

Notice the sentence "if implemented as written above, you get algebraic loops". This means that many DAEs can be rewritten into a set of ODEs, but this can be a pretty complex math challenge and I am not going into details here.