how to design a function to capture the boundary condition for $u(x=-1,t)=u(x=1,t)$ (for a PDE)?

27 Views Asked by At

I need to come up with a boundary condition function for a PDE solving method desribed here: https://github.com/sciann/sciann-applications/blob/master/SciANN-BurgersEquation/SciANN-BurgersEquation.ipynb

The condition is $$u(x=-1,t)=u(x=1,t)$$ The way I see it I need to come up with two expression:

  • Expression 1: $g_1(x) * u$ st. $g =0$ when $x\neq -1$ and $g=1$ for $x=-1$
  • Expression 2: $g_2(x) * u$ st. $g =0$ when $x\neq1$ and $g=1$ for $x=1$

Then my BC condition function $f$ is $$f=g_1u-g_2u$$ which in solving process is set to $f=0$. How should I formulate this function? I probably need some sigmund function but, right? Please help