Solving an integral when the parameters is a matrix - Controllability Gramian
2 visualizaciones (últimos 30 días)
Mostrar comentarios más antiguos
Hi,
I need solving the controllability gramian below given B and psi
Psi is calculated as below as phi_s. B matrix is give below. [to tf]=[0 10]
clear all;clc;close all
syms t;
A = [0 1 0 0;0 0 -4.91 0;0 0 0 1;0 0 73.55 0];
B = [0;2;0;-10];
[M,J] = jordan(A);
phi_s=M * expm(J*t) * inv(M);
figure(3)
fplot([phi_s(1,1) phi_s(1,2) phi_s(1,3) phi_s(1,4) ...
phi_s(2,1) phi_s(2,2) phi_s(2,3) phi_s(2,4) ...
phi_s(3,1) phi_s(3,2) phi_s(3,3) phi_s(3,4) ...
phi_s(4,1) phi_s(4,2) phi_s(4,3) phi_s(4,4)],[0 2])
0 comentarios
Respuesta aceptada
Torsten
el 28 de Nov. de 2022
You mean
syms t t0 tf
A = [0 1 0 0;0 0 -4.91 0;0 0 0 1;0 0 73.55 0];
B = [0;2;0;-10];
[M,J] = jordan(A);
phi_s=M * expm(J*t) * inv(M);
Mat = (phi_s*B) * (phi_s*B).';
G_c = int(Mat,t0,tf)
size(G_c)
?
3 comentarios
Torsten
el 28 de Nov. de 2022
It's done automatically if you leave away the ";" after the line of the generated symbolic expression.
Paul
el 29 de Nov. de 2022
Torsten's solution by direct, symbolic integration
syms t t0 tf real
assumeAlso(tf >= t0);
A = sym([0 1 0 0;0 0 -4.91 0;0 0 0 1;0 0 73.55 0]);
B = sym([0;2;0;-10]);
[M,J] = jordan(A);
phi_s=M * expm(J*t) * inv(M);
Mat = (phi_s*B) * (phi_s*B).';
G_c(t0,tf) = int(Mat,t0,tf)
Sheng's solution using symbolic expm
S(t0,tf) = G_c1(t0,tf,A,B);
Show they are equivalent
E = simplify(G_c - S)
The latter can, in principle, also be used for direct numerical evaluation, (assuming no overflows, etc.), compare to vpa of the symbolic result
vpa(S(1,2.5))
format long e
G_c1(1,2.5,double(A),double(B))
function out = G_c1(t0,tf,A,B)
Qs = expm(A*t0)*B;
Q = Qs*Qs.';
temp = (tf-t0)*[-A Q; 0*A A.'];
temp = expm(temp);
if isa(temp,'sym')
out = simplify(expand(simplify(expand(temp(end/2+1:end,end/2+1:end).')) * simplify(expand(temp(1:end/2,end/2+1:end)))));
else
out = temp(end/2+1:end,end/2+1:end).' * temp(1:end/2,end/2+1:end);
end
end
Más respuestas (0)
Ver también
Categorías
Más información sobre Linear Algebra en Help Center y File Exchange.
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!