e.g. y~0 = A~y ! a11 a12 where A = and ~y (t) = a21 a22
y1 (t)
y2 (t)
Characteristic equation of planar system: 2 T + D = 0
Eigenvalues of A: = (T
T 2 4D)/2 where T = tr(A) and D = det(A)
1. T2 4D > 0: 2 distinct real roots
2. T2 4D < 0: 2 complex conjugate roots 3. T2 4D = 0: 1 real root of multiplicity 2 Proposition 2.4: Let 1 and 2 be eigenvalues of A, and v~1 and v~2 be their respective nonzero eigenvectors. If 1 6= 2 , then v~1 and v~2 are linearly independent. Corollary 2.8: If 1 6= 2 , then the solutions y~1 (t) = e1 t v~1 and y~2 (t) = e2 t v~2 are linearly independent on R. Note: Both are true for cases 1 and 2.
(1) 2 distinct real roots
~ = C1 y~1 (t) + C2 y~2 (t) y(t) = C2 e1 t v~1 + C2 e2 t v~2
(1)
(2) 2 complex conjugate roots
~ = C1 et w y(t) ~ + C2 et w ~
= C1 et (costv~1 sintv~2 ) + C2 et (sintv~1 + costv~2 )
(3) 1 real root of multiplicity 2 1. Find eigenvalues and eigenvectors. 1
(2)
2. If eigenspace = 1, treat as normal. If eigenspace = 2, continue.
3. Form y~1 (t) = et v~1 4. Find a v~2 such that (A I)v~2 = v~1 (a) Use w ~ to test (any vector independent of v~1 would do) (b) Compute (A I)w ~ (c) Adjust w ~ accordingly 5. Form y~2 (t) = et (v~2 + tv~1 ) 6. General solution: ~y (t) = C1 y~1 + C2 y~2 = et [(C1 + C2 t)v~1 + C2 v~2 ]