how to solve the following equation to get lambda values from it?

5 次查看(过去 30 天)
modulus([2 -1; -1 2]-λ[1 0; 0 1])=0
please help me. thanks.

回答(1 个)

Walter Roberson
Walter Roberson 2020-8-4
编辑:Walter Roberson 2020-8-5
modulus is a matrix operator; on a 2 x 2 input, it would give a 2 x 2 output. You ask that it be equal to 0: are you asking that it equal [0 0; 0 0]?
If so then because the modulus of a matrix A is X such that X*X = A'A and X would be [0 0; 0 0] then we can see that A'A would have to equal 0.
syms lambda
A = [2 -1; -1 2]-lambda * [1 0; 0 1]
B = simplify(A'*A)
B =
[ 1 - ((2*lambda - abs(lambda)^2)*(lambda - 2))/lambda, 2*real(lambda) - 4]
[ 2*real(lambda) - 4, 1 - ((2*lambda - abs(lambda)^2)*(lambda - 2))/lambda]
B(2,1) and B(1,2) are 2*real(lambda)-4 and that has to equal 0, which requires that lambda = 2 + 1i*LI where LI is real. Substituting that in,
syms LI real
D = subs(B,lambda,2+1i*LI)
The diagonal becomes the same, 1 - (LI*(- LI^2 + LI*2i)*1i)/(2 + LI*1i) . Normalize it,
>> simplify(D(1)*(2+LI*1i))
ans =
LI^3*1i + 2*LI^2 + LI*1i + 2
and that has to equal 0, and by construction LI is real valued. Looking at the real part, 2*LI^2 + 2 = 0... which is only possible for imaginary LI, but by construction LI is real. LI^3 = -LI would also have to be true to take care of the imaginary components, but that is not possible for real-valued LI.
We have arrived at a contradiction, which can potentially be resolved in one of two ways:
  1. That no such modulus exists; or
  2. That the normalization (2+LI*1i) was invalid because (2+LI*1i) = 0. But for (2+LI*1i) = 0, LI would have to be imaginary, which again leads to a contradiction.
So, there is no solution.

类别

Help CenterFile Exchange 中查找有关 Ordinary Differential Equations 的更多信息

Community Treasure Hunt

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

Start Hunting!

Translated by