Bvp4c: Unable to solve the collocation equations -- a singular Jacobian encountered.

Question

Marcus Rosales 2022-4-6

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1690190-bvp4c-unable-to-solve-the-collocation-equations-a-singular-jacobian-encountered

回答： Marcus Rosales 2022-4-10

I am trying to solve a boundary value problem, but I keep getting the above error. I know there are similar posts to this, but I can not seem to get this to work with them, so maybe someone can look directly at my code to get an answer?

The differential equation I want to solve is (putting it in first order form and wrting as a row vector):

. Here:

I included my code as well and used

to pick an inital guess. This is doing a bad job of that I think however! What else could I do here? Another inital guess, another function? Any help will be appreciated!

2 个评论
显示无隐藏无

Torsten 2022-4-6

I mean to remember that this is a transformed equation with tanh to map (-oo,oo) to [-1:1].

Could you include the original equation as well ?

Marcus Rosales 2022-4-6

I updated the attached code, and I sure can give the original equation:

In retrospect I should have included this anyways. Mapping the intreval how I did may not be the best route... Maybe someone can see something better from here.

I simply tried settting the boundary conditions to a big number for the equation in this form, but got something oscillatory or a singular Jacobian (as well).

Also, these expressions are equivalent:

so the divergence of the jacobian is at $x=\pm 1$. Restricting my interval I was able to find solutions, but with huge residuals! Not sure if something is wrong with the code itself, or the approach.

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Marcus Rosales 2022-4-10

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1690190-bvp4c-unable-to-solve-the-collocation-equations-a-singular-jacobian-encountered#answer_939310

I guess I'll answer this question since I found it out.

It turned out to work better using the EOM for α. The main issue here is as t blows up, the derivative (after converting to first order system of equations) begins to blow up making various RK-methods unstable.

We first can try fixing the b.c. at something like

for l big but it turns out, here, to model the profile (amplitude) I want to. It turns out setting the upper limit to somehting just under 1 works well, then stretch the solution for

by a tanh.