MA1201 Mathematics-II

Tutorial sheet-3
System of Linear Differential Equations
-Rough Ideas of some solutions (may also be incorrect). First Check it,
before you read it thoroughly

1. Find the general solution of the following Cauchy-Euler differential equations:

(a) Prove that the transformation x = et reduces the equation
a0 x2 y ′′ + a1 xy ′ + a2 y = F (x) (Cauchy Euler Equation)
to a linear differential equation with constant coefficients.
(b) x2 y ′′ − 4xy ′ + 4y = 4x2 − 6x3 , y(2) = 4, y ′ (2) = −1;
(c) (2x − 3)2 y ′′ − 6(2x − 3)y ′ + 12y = 0
2. The equation
dy
= A(x)y 2 + B(x)y + C(x), (1)
dx
is called Riccati’s equation.
(a) Solve the following differential equations
dy
dx
= 2y 2 − 3y + 1, y1 (x) = 1
dy
dx
= 2xy 2 − 2x2 y + 1, y1 (x) = x

3. Suppose that the roots of the characteristic equation of the system

x′ = a1 x + b1 y
(2)
y ′ = a2 x + b2 y
are real and equal; and let λ denote their common value. Also assume that the
given system is not such that a1 = b2 ̸= 0 and a2 = b1 = 0. Then show that there
exist no nontrivial solutions of the form x = Ateλt , y = Bteλt , which is linearly
independent of the "basic" solution of the form x = Aeλt , y = Beλt .
4. Find a unique solution ϕ of the given differential equation
   
7 −1 6 x1
dx 
= −10 4 −12 x, where x = x2  ,
 
dt
−2 1 −1 x3
that satisfies the initial condition
 
−1
ϕ(0) = x0 , where x0 =  4  .
2

5. Consider the non-homogeneous differential equation

   5t 
dx 6 −3 e
= x+ ,
dt 2 1 4
Here, x is a vector. Find the general solution.
 6. Find the critical points, its type, and the general solutions of the following systems:
(
dx
dt
= −3x + 4y,
(a) dy
dt
= −2x + 3y,
(
dx
dt
= −4x − y,
(a) dy
dt
= x − 2y,
(
dx
dt
= x − 2y,
(a) dy
dt
= 4x + 5y.

7. Consider the following non-linear systems

dx dx
(A) = sin x − 4y (B) = 8x − y 2
dt dt
dy dy
= sin 2x − 5y = −6y + 6x2
dt dt
For each system, determine the nature and stability of their critical points.

Solutions
1. These have been covered in class.
2. (a) Solution:
dy
= 2y 2 − 3y + 1, y1 (x) = 1
dx
Step 1: Find a particular solution y1
We are given that y1 (x) = 1 is a particular solution.
Step 2: Perform a substitution.
Let’s introduce a new variable ‘z’ such that:

1 1
= 1 + , is the solution.
y = y1 +
z z
Step 3: Differentiate y with respect to x

dy 1 dz
=− 2
dx z dx
Step 4: Substitute y and dy/dx into the original equation
 2  
1 dz 1 1
− 2 =2 1+ −3 1+ +1
z dx z z
Step 5: Simplify the equation
 1 dz 4 2 3
− = 2 + + − 3 − +1
z 2 dx z z2 z
1 dz 1 2
− 2 = + 2
z dx z z
dz
− =2+z
dx
Step 6: Solve the resulting first-order linear differential equation

dz
+ dx = 0, loge (2 + z) = C − x, z = Ae−x − 2.
2+z
Step 7: Substitute z back into the expression for y, we get
1
The final answer is , y = 1 +
Ae−x − 2
(b) Solution:
dy
= 2xy 2 − 2x2 y + 1, y1 (x) = x
dx
Step 1: Find a particular solution y1
We are given that y1 (x) = x is a particular solution.
Step 2: Perform a substitution.
Let’s introduce a new variable ‘z‘ such that:

1 1
y = y1 + =x+ (3)
z z
Step 3: Differentiate y with respect to x

dy 1 dz
=1− 2
dx z dx
Step 4: Substitute y and dy/dx into the original equation
 2  
1 dz 1 2 1
1− 2 = 2x x + − 2x x + +1
z dx z z
Step 5: Simplify the equation
   
1 dz 1 1
1− 2 = 2x x + x+ −x +1
z dx z z
  
1 dz 1 1
1− 2 = 2x x + +1
z dx z z
1 dz
− 2 = 2x (zx + 1)
z dx
dz
− = 2x + 2x2 z
dx
dz
+ 2x2 z = −2x
dx
Step 6: Solve the integral; the integrating factor, u(x), is given by:
2x2 dx) 3
R
u(x) = e( = e(2x /3)
(4)
Step 7: From here, we calculate the z
Z
3 3
ze(2x /3) = −2xe(2x /3) dx (5)

Step 8: Now, eliminate z from (3) and (5), we get the answer.
Question 3: Suppose there exists a non-trivial solution of the form
x = Ateλt , y = Bteλt . (6)
If (6) is a solution of the given system, then it must satisfy the equations
of the system. So, differentiating (6) and substituting into the system, we
must have
(Aλt + A)eλt = a1 Ateλt + b1 Bteλt ,
(Bλt + B)eλt = a2 Ateλt + b2 Bteλt .
These equations reduce to
[(a1 − λ)A + b1 B]t − A =0,
[(b2 − λ)B + a2 A]t − B =0.
This implies
[(a1 − λ)A + b1 B] = 0, A = 0;
[(b2 − λ)B + a2 A] = 0, B = 0.
Hence, we must have A = B = 0, which means that (6) must be the trivial
solution. This is a contradiction. Thus our supposition is invalid and there
exists no nontrivial solution of the stated form.
 Question 4 Suppose we seek the unique solution ϕ of the differential
equation
   
7 −1 6 x1
dx 
= −10 4 −12 x, where x = x2  ,
dt
−2 1 −1 x3
(11.10) (7)
that satisfies the initial condition
 
−1
ϕ(0) = x0 , where x0 =  4  .
2
(11.28) (8)
By Theorem 11.10, this solution is given by

ϕ(t) = Φ(t)Φ−1 (0)x0 ,

(11.29) (9)
where Φ(t) is a fundamental matrix of differential equation (11.10). In
Example 11.9 (SL Ross) we observed that
 2t
e3t 3e5t

e
Φ(t) = −e2t −2e3t −6e5t 
−e2t −e3t −2e5t
is such a fundamental matrix. After performing the required calculations,
we find
 −2t
e−2t

2e 0
Φ−1 (t) = −4e−3t −e−3t −3e−3t  ,
e−5t 0 e−5t
and hence
 
2 1 0
Φ−1 (0) = −4 −1 −3 .
1 0 1
Thus, using (11.29), the solution of the problem is
 2t
e3t 3e5t
  
e 2 1 0 −1
2t 3t 5t
Φ(t) = −e −2e −6e  −4 −1 −3  4 
−e2t −e3t −2e5t 1 0 1 2
 e2t e3t 3e5t
  
2
= −e2t −2e3t −6e5t  −6
−e2t −e3t −2e5t 1
2e2t − 6e3t + 3e5t
 

= −2e2t + 12e3t − 6e5t  .

−2e2t + 6e3t − 2e5t
We write this as

x1 = 2e2t − 6e3t + 3e5t ,

x2 = −2e2t + 12e3t − 6e5t ,

x3 = −2e2t + 6e3t − 2e5t .

(11.30) (10)
We should note that the determination of Φ−1 (t) requires some calculations
(we again refer to the aforementioned Exercise 1(a)). For this reason, the
following alternate procedure for solving the problem is preferable.
In Example 11.8, we noted that the general solution of the differential
equation (11.10) can be written as

x1 = c1 e2t + c2 e3t + 3c3 e5t ,

x2 = −c1 e2t − 2c2 e3t − 6c3 e5t ,

x3 = −c1 e2t − c2 e3t − 2c3 e5t .

(11.31) (11)
In terms of components, the given initial condition (11.28) is

x1 (0) = −1, x2 (0) = 4, x3 (0) = 2.

Applying these conditions to (11.31), we at once have

c1 + c2 + 3c3 = −1,

−c1 − 2c2 − 6c3 = 4,

 −c1 − c2 − 2c3 = 2.

(11.32) (12)
This system of three algebraic equations in three unknowns has the unique
solution c1 = 2, c2 = −6, c3 = 1.
Question 5: Consider the nonhomogeneous differential equation
   5t 
dx 6 −3 e
= x+
dt 2 1 4
Where,

e5t
     
6 −3 x1
A(t) = , F(t) = , and x = .
2 1 4 x2
The corresponding homogeneous differential equation is
 
dx 6 −3
= x
dt 2 1
We found that a general solution to the system,
 3t   4t 
e 3e
ϕ1 (t) = 3t and ϕ2 (t) =
e 2e4t
Constitute a fundamental set (pair of linearly independent solutions).
Thus, a fundamental matrix is given by
 3t
e 3e4t

Φ(t) =
e3t 2e4t
We know that,
Z 1
ϕ0 (t) = Φ(t) Φ−1 (u)F(u)du
t0
Is a solution of a considered nonhomogeneous differential equation for
every real number t0 . For convenience, we choose t0 = 0. We thus proceed
to evaluate
Z 1
ϕ0 (t) = Φ(t) Φ−1 (u)F(u)du
0
3t 4t
 
e 3e e5u


putting Φ(t) = and F (u) = . We find that
e3t 2e4t 4
 −2e−3u 3e−3u
 
−1
Φ (u) =
e−4u −e−4e
Then,
Z t   5u 
e3t 3e4t −2e−3u 3e−3u

e
ϕ0 (t) = du
e3t 2e4t 0 e−4u −e−4u 4
Z 1
−2e2u + 12e−3u
 3t
e 3e4t

= du
e3t 2e4t 0 eu − 4e−4u
−e − 4e−3t + 5
 3t   2t
e 3e4t

=
e3t 2e4t et + e−4t − 2
 5t
2e − 1 + 5e3t − 6e4t

=
e5t − 2 + 5e3t − 4e4t
This, then, is a solution of the nonhomogencous differential equation. In
component form, it is

x1 = 5e3t − 6e4t + 2e5t − 1

x2 = 5e3t − 4e4t + e5t − 2
Question 6:
(a) The system is (
dx
dt = −3x + 4y,
dy
dt = −2x + 3y.
dx dy
Critical Point: Setting dt = 0 and dt = 0, we get
−3x + 4y = 0,
−2x + 3y = 0.
Solving, we find the critical point is (x, y) = (0, 0).
Type of Critical Point: The coefficient matrix is
 
−3 4
A= .
−2 3
we find its eigenvalues by solving the characteristic equation:
det(A − λI) = 0,
where I is the 2 × 2 identity matrix.
     
−3 4 1 0 −3 − λ 4
A − λI = −λ = .
−2 3 0 1 −2 3−λ
 det(A − λI) = (−3 − λ)(3 − λ) − (4)(−2).
det(A − λI) = (−9 − 3λ + 3λ + λ2 ) + 8 = λ2 − 1.
Now,
λ2 − 1 = 0.
The eigenvalues of A are:
λ1 = 1 and λ2 = −1.
eigenvalues are real and of opposite sign so type of critical point is
Saddle Point.
Now, we compute the eigenvectors corresponding to these eigenvalues.
Eigenvector for λ1
Substitute λ = 1 into A − λI:
     
−3 4 1 0 −4 4
A − λI = − = .
−2 3 0 1 −2 2

To find the eigenvector, solve:

    
−4 4 x1 0
= .
−2 2 x2 0

This simplifies to the equation:

−4x1 + 4x2 = 0 =⇒ x1 = x2 .

Let x2 = t. Then x1 = t,
  
t 1
v1 = =t .
t 1

Thus, the eigenvector corresponding to λ1 is:

 
1
1
Eigenvector for λ2 ,Substitute λ = −1 into A − λI:
     
−3 4 −1 0 −2 4
A − λI = − = .
−2 3 0 −1 −2 4

To find the eigenvector, solve:

    
−2 4 x1 0
= .
−2 4 x2 0
 This simplifies to the equation:

−2x1 + 4x2 = 0 =⇒ x1 = 2x2 .

Let x2 = t. Then x1 = 2t,

   
2t 2
v2 = =t .
t 1

Thus, the eigenvector corresponding to λ2 is:

 
2
1

General Solution: The solution is of the form

   
1 t 2 −t
X(t) = C1 e + C2 e ,
1 1
where C1 and C2 are constants.
(b) The system is (
dx
dt = −4x − y,
dy
dt = x − 2y.
dx dy
Critical Point: Setting dt = 0 and dt = 0, we get

−4x − y = 0,
x − 2y = 0.
Solving, we find the critical point is (x, y) = (0, 0).
Type of Critical Point: The coefficient matrix is
 
−4 −1
A= .
1 −2
We find its eigenvalues by solving the characteristic equation:

det(A − λI) = 0,

where I is the 2 × 2 identity matrix.

     
−4 −1 1 0 −4 − λ −1
A − λI = −λ = .
1 −2 0 1 1 −2 − λ

det(A − λI) = (−4 − λ)(−2 − λ) − (−1)(1).

 det(A − λI) = (λ + 4)(λ + 2) − (−1).

det(A − λI) = λ2 + 6λ + 9.

det(A − λI) = (λ + 3)2 .

The eigenvalues of A are:
λ1 = λ2 = −3.

Since both eigenvalues are the same and real so type of critical point
is Node.
Eigenvector for λ = −3: Substitute λ = −3 into A − λI:
     
−4 −1 −3 0 −1 −1
A − λI = − = .
1 −2 0 −3 1 1

To find the eigenvector, solve:

    
−1 −1 x1 0
= .
1 1 x2 0

This simplifies to:

−x1 − x2 = 0 =⇒ x1 = −x2 .

Let x2 = t. Then x1 = −t,

   
−t −1
v1 = =t .
t 1

Thus, the eigenvector corresponding to λ = −3 is:

 
−1
v1 = .
1

Generalized Eigenvector for λ = −3: To find the generalized eigenvec-

tor, solve:
(A − λI)v2 = v1 .
Substituting λ = −3, we solve:
    
−1 −1 x1 −1
= .
1 1 x2 1
 This simplifies to:

−x1 − x2 = −1, x1 + x2 = 1.

Solving these equations, we find x1 = 0 and x2 = 1. Thus, the gener-

alized eigenvector is:  
0
v2 = .
1
General Solution: The solution is of the form:
      
−1 −3t 0 −3t −1 −3t
X(t) = C1 e + C2 e + te ,
1 1 1
where C1 and C2 are constants.
(c) The system is (
dx
dt = x − 2y,
dy
dt = 4x + 5y.
dx dy
Critical Point: Setting dt = 0 and dt = 0, we get

x − 2y = 0,
4x + 5y = 0.

Solving, we find the critical point is (x, y) = (0, 0).

Type of Critical Point: The coefficient matrix is
 
1 −2
A= .
4 5
We find its eigenvalues by solving the characteristic equation:

det(A − λI) = 0,

where I is the 2 × 2 identity matrix.

     
1 −2 1 0 1 − λ −2
A − λI = −λ = .
4 5 0 1 4 5−λ

det(A − λI) = (1 − λ)(5 − λ) − (−2)(4).

det(A − λI) = (1 − λ)(5 − λ) + 8 = λ2 − 6λ + 13.

The eigenvalues of A are the roots of the quadratic equation:
λ2 − 6λ + 13 = 0.

we find: √ √
6± 36 − 52 6 ± −16
λ= = = 3 ± 2i.
2 2
The eigenvalues are:
λ1 = 3 + 2i, λ2 = 3 − 2i.
nature of eigenvalues are complex conjugate but not pure imaginary so
type of critical point is Spiral.
Eigenvectors: For λ = 3 + 2i, substitute into A − λI:
     
1 −2 1 0 −2 − 2i −2
A − λI = − (3 + 2i) = .
4 5 0 1 4 2 − 2i
To find the eigenvector, solve:
    
−2 − 2i −2 x1 0
= .
4 2 − 2i x2 0
From the first row:
2 2
(−2 − 2i)x1 − 2x2 = 0 =⇒ x1 = x2 = x2 .
−2 − 2i −(2 + 2i)
Simplify:
2 2 − 2i 4 − 4i i − 1
· = = .
−(2 + 2i) 2 − 2i −8 2
Thus:
i−1
x1 = x2 .
2
Thus eigen vector respect to the value 3 + 2i is:
 i−1 
v1 = 2 .
1
 i−1  1 
(3+2i)t 2 3t 2 [− cos(2t) − sin(2t) + i(cos(2t) − sin(2t))]
e =e .
1 cos(2t) + i sin(2t)
 − cos(2t)−sin(2t)   cos(2t)−sin(2t) 
Xr (t) = e3t 2 , Xi (t) = e3t 2 .
cos(2t) sin(2t)
General Solution: The solution is of the form:
X(t) = C1 Xr (t) + C2 Xi (t)

Here, C1 and C2 are arbitrary constants.