Practice Problems for Section 7.9: Solutions cos 2t 1 0 2 ] 1 Solve x0 = Ax + , x(0) = , A = . − sin 2t 0 −2 0 ¯ ¯ ¯ −r 2 (i) Find eAt : |A − rI| = ¯¯ ¯ −2 −r ¯ ¯ ¯ ¯ = r 2 + 4 = 0 ⇒ r = ±2i. ¯ ¯ −2i 2 1 1 0 A − 2iI = ⇒ eigenvector ξ = = + i = a + ib. −2 −2i i 0 1 cos 2t sin 2t u(t) = a cos 2t − b sin 2t = , v(t) = a sin 2t + b cos 2t = . − sin 2t cos 2t cos 2t sin 2t Hence Ψ(t) = [u(t), v(t)] = . Since Ψ(0) = I ⇒ Ψ(t) = eAt . − sin 2t cos 2t cos 2t sin 2t 1 cos 2t (ii) Homogeneous part of solution: eAt x0 = = . − sin 2t cos 2t 0 − sin 2t (iii) Nonhomogeneous part of solution: R t cos 2s − sin 2s cos 2s 1 ⇒ 0t e−As f (s)ds = , = e−As f (s) = 0 0 − sin 2s sin 2s cos 2s R ⇒ eAt 0t e−As f (s)ds = cos 2t sin 2t t t cos 2t . = −t sin 2t 0 − sin 2t cos 2t (iv) Full solution: At At x(t) = e x0 + e Z t 0 −As e cos 2t f (s)ds = (1 + t) . − sin 2t 2t cos 2t 1 0 2 . ] 2 Solve x0 = Ax + , x(0) = , A = −2 0 −2t sin 2t 0 cos 2t cos 2t sin 2t At , e x0 = . A and x0 are as in ] 1 ⇒ eAt = − sin 2t cos 2t − sin 2t Compute nonhomogeneous part: R cos 2s − sin 2s 2s cos 2s 2s t2 , e−As f (s) = = ⇒ 0t e−As f (s)ds = 0 sin 2s cos 2s −2s sin 2s 0 R t ⇒ eAt 0 e−As f (s)ds = cos 2t sin 2t t2 t2 cos 2t = . 0 −t2 sin 2t − sin 2t cos 2t ⇒ x(t) = eAt x0 + eAt Z t 0 cos 2t e−As f (s)ds = (1 + t2 ) . − sin 2t 1 36t2 0 2 −4 ] 3 Solve x0 = Ax + , x(0) = , A = . 6t 1 1 −2 ¯ ¯ ¯ 2−r At (i) Find e : |A − rI| = ¯¯ ¯ 1 Verify A2 = −4 −2 − r ¯ ¯ ¯ ¯ = r 2 = 0 ⇒ double eva r = 0. ¯ ¯ 0 0 1 + 2t −4t ⇒ eAt = I + At = . 0 0 t 1 − 2t (ii) Homogeneous part of solution: eAt x0 = (iii) Nonhomogeneous part of solution: 1 + 2t −4t 0 −4t = . t 1 − 2t 1 1 − 2t 1 − 2s 4s 36s2 60s2 − 72s3 e−As f (s) = = −s 1 + 2s 6s 6s + 12s2 − 36s3 R t −As ⇒ 0 e f (s)ds = 20t3 − 18t4 3t2 + 4t3 − 9t4 3 4 3 4 2 3 4 R 8t + 6t (1 + 2t)(20t − 18t ) − 4t(3t + 4t − 9t ) t = ⇒ eAt 0 e−As f (s)ds = 3t2 − 2t3 + 3t4 t(20t3 − 18t4 ) + (1 − 2t)(3t2 + 4t3 − 9t4 ) (iv) Final solution: x(t) = eAt x0 + eAt Z t 0 e−As f (s)ds = 3 (i) Find eAt : Matrix is upper triangular ⇒ triple eva r = 1. 0 2 3 0 0 4 2 3 Set B = A − I = 0 0 2 ⇒ B = 0 0 0 , B = 0, 0 0 0 0 0 0 hence eAt = et eBt 1 2t 3t + 2t2 . = et (I + Bt + B 2 t2 /2) = et 2t 0 1 0 0 1 (ii) Compute solution: 1 −2s −3s + 2s2 0 −18s + 12s2 0 = e−As f (s) = e−s 1 −2s −12s 0 s 0 0 1 6 6e 2 −4t + 8t + 6t 1 − 2t + 3t2 − 2t3 + 3t4 0 0 1 2 3 0 ] 4 Solve x = Ax + 0 , x(0) = 0 , A = 0 1 2 . 6et 0 0 0 1 4 −9t2 + 4t3 R t −As , hence ⇒ 0 e f (s)ds = −6t2 6t x(t) = eAt Z t 0 1 2t 3t + 2t2 −9t2 + 4t3 e−As f (s)ds = et 2t −6t2 0 1 0 0 1 6t (−9t2 + 4t3 ) − 12t3 + 6t(3t + 2t2 ) 9t2 + 4t3 = et = et 6t2 −6t2 + 12t2 6t 6t 100t 0 6 −7 ] 5 Solve x0 = Ax + , x(0) = , A = . −50t 0 1 −2 >> syms t s;A=sym([6 -7;1 -2]);f=[100*s;-50*s]; >> simplify(int(expm(-A*s)*f,0,t)); >> x=simplify(expm(A*t)*ans) x = [ -exp(5*t)*(110*t*exp(-5*t)-68*exp(-5*t)-7+75*exp(-6*t))] [ -exp(5*t)*(80*t*exp(-5*t)-74*exp(-5*t)+75*exp(-6*t)-1)] 2 −5 0 4t . ] 6 Solve x0 = Ax + , x(0) = , A = 1 0 1 −2 >> syms t s;A=sym([2 -5;1 -2]);f=[4*s;1]; >> simplify(int(expm(-A*s)*f,0,t)); >> x=simplify(expm(A*t)*ans) x = [ cos(t)-8*sin(t)+8*t-1] [ -3*sin(t)+2*cos(t)+4*t-2] 4 cos t 3 2 −5 ] 7 Solve x0 = Ax + , x(0) = , A = . 6 sin t 5 1 −2 3 >> syms t s;A=sym([2 -5;1 -2]);f=[4*cos(s);6*sin(s)]; >> simplify(int(expm(-A*s)*f,0,t)); >> x=simplify(expm(A*t)*([3;5]+ans)) x = [ 3*cos(t)+17*t*cos(t)-32*sin(t)+4*sin(t)*t] [ -13*sin(t)+5*sin(t)*t+5*cos(t)+6*t*cos(t)] e−t cos 2t 0 −1 0 0 0 −t ] 8 Solve x = Ax + e sin 2t , x(0) = 0 , A = 3 1 −2 . 0 0 3 2 2 >> syms t s;A=sym([-1 0 0;3 1 -2;3 2 1]),f=[exp(-s)*cos(2*s);exp(-s)*sin(2*s);0]; >> x=simplify(factor(simple(expm(A*t)*int(expm(-A*s)*f,0,t)))) x = [ 1/2*sin(2*t)*exp(-t)] [ -1/20*(18*exp(-2*t)*sin(2*t)+11*exp(-2*t)*cos(2*t)-11*cos(2*t)-7*sin(2*t))*exp(t)] [ -1/20*exp(t)*(4*exp(-2*t)*sin(2*t)-7*exp(-2*t)*cos(2*t)-11*sin(2*t)+7*cos(2*t))] ]9 0 2 −1 (a) Find a particular solution to x0 = Ax + t , where A = , 1 3 −2 via the method of undetermined coefficients. Eigenvalues of A: (2 − r)(−2 − r) + 3 = r2 − 1 = 0 ⇒ r = ±1. Since 0 is not an eigenvalue, the form of the particular solution is xp (t) = at + b. Sub this in ODE system: 0 0 x0p = a = Axp + t = Aat + Ab + t . 1 1 From coefficients of t0 ⇒ a = Ab. 0 0 2 −1 0 1 From coefficients of t1 ⇒ 0 = Aa + ⇒ a = −A−1 = − = . 1 1 3 −2 1 2 4 2 −1 1 0 Then b = A−1 a = = , hence 3 −2 2 −1 t . xp (t) = at + b = 2t − 1 1 2 −1 (b) Find a particular solution to x0 = Ax + et , where A = , 0 3 −2 via the method of undetermined coefficients. Matrix is the same as in (a). Since 1 is an eigenvalue of A, the form of the particular solution is xp (t) = et (at + b). Sub this in ODE system: 1 1 x0p = et (at + a + b) = Axp + et = et (Aat + Ab + ). 0 0 From coefficients of t1 et ⇒ a = Aa (∗) 1 1 From coefficients of t0 et ⇒ a + b = Ab + ⇒ (A − I)b = a − (∗∗) 0 0 1 (∗) ⇒ a is an eigenvector of A for r = 1. It is easy to see that is such an eigenvector, 1 1 hence a = ξ with ξ to be determined. An equation for ξ follows from the requirement that 0 1 (∗∗) has a solution. This is equivalent to yT (a − ) = 0 for any solution of (A − I)T y = 0. 0 3 An eigenvector of AT for the eigenvalue 1 is y = , thus the condition for (∗∗) to have a −1 solution is ξ 1 3/2 1 1/2 [3, −1]( − ) = 2ξ − 3 = 0 ⇒ ξ = 3/2 ⇒ a = and a − = . ξ 0 3/2 0 3/2 Since it is now that (∗∗) has a solution, we only need to evaluate the first component. guaranteed 1/2 b 1 , this yields b1 − b2 = 1/2. We choose b2 = 0 ⇒ b1 = 1/2, hence b = . Writing b = 0 b2 The particular solution then becomes et 3t + 1 xp (t) = et (at + b) = . 2 3t 5