Finite difference method 1st degree

3 vues (au cours des 30 derniers jours)
Patirick Legare
Patirick Legare le 7 Déc 2019
Commenté : J. Alex Lee le 7 Déc 2019
I have to solve this problem using a first degree finite difference method. I am really struggling to understand what is not working with my code. If someone could take a look at ir and explain to me what<s wrong I would really appreciate it. Here is the assignement and the code below Untitled.jpg
Untitled2.jpg
function[y]=problimite(N,P,Q,R, a, b, alpha, beta)
h=(b-a)/(N+1);
S=-1+P(1:end-1)*h/2;
D=2+Q(1:end)*h^2;
I=-1-P(2:end)*h/2
B=[(-R(1)*(h^2)+(1+P(1)*h/2))*alpha,-R(2:end-1).*h^2,-R(end)*(h^2)+(1+P(end)*h/2)*beta];
disp(S)
y = tridiagonal(N, S, D, I, B);
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%5
function y = tridiagonal(N, S, D, I, B)
y(1)=B(1)/D(1);
for i=2:N
S(i-1)=(S(i-1))/D(i-1);
D(i)=D(i)-I(i-1)*S(i-1);
y(i)=(B(i)-I(i-1)*y(i-1))/D(i);
end
x(N)=y(N);
for i=N-1:1
x(i)=y(i)-S(i)*x(i+1);
end
%%script%%%%
b=1;
alpha=0;
beta=1;
X=[0.9:0.01:1];
p=@(x) -1/x;
q=@(x) 0;
r=@(x) (1.6)/(x^4);
N=9;
P=arrayfun(p,X);
R=arrayfun(r,X);
Q=arrayfun(q,X);
%h=.01;
%I=[-1-P(2:end)*h/2]
%D=[2+Q(1:end)*h^2]
%S=[-1+P(1:end-1)*h/2]
%display(P);
%display(R);
%display(Q);
plot(problimite(N,P,Q,R, a, b, alpha, beta))
hold on
%% partie 2 %%
X=[0.9:0.005:1];
p=@(x) -1/x;
q=@(x) 0;
r=@(x) (1.6)/(x^4);
N=19;
P=arrayfun(p,X);
R=arrayfun(r,X);
Q=arrayfun(q,X);
%display(P);
%display(R);
%display(Q);
plot(problimite(N,P,Q,R, a, b, alpha, beta))
  2 commentaires
darova
darova le 7 Déc 2019
Can't you use bvp4c? It can solve this task
J. Alex Lee
J. Alex Lee le 7 Déc 2019
@darova, this looks like a homework assignment requiring hand-coding; it even appears to have an analytical solution for verification.
@Patirick, what do you mean by "first degree" finite difference method? Do you mean first order accuracy? If so I'm not sure there is a first order accurate difference formula for second order derivatives...You can break up your problem into a pair of coupled first order ODEs and use a pair of first order accurate finite differencing schemes. Perhaps you can explain in words the strategy that your code is implementing before jumping into the coding.

Connectez-vous pour commenter.

Réponses (0)

Catégories

En savoir plus sur Dynamic System Models dans Help Center et File Exchange

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by