Chapter2

SYSTEMS OF LINEAR DIFFERENTIAL

EQUATIONS

Linear differential equations are obtained in modeling various applications such as electrical
circuits, vibration systems mixture of substances etc. Consider for example a mass and spring
system in the Figure 2.1.

Figure 2.1 Modelling a spring and mass system.

Newton’s 2nd law states that; ∑F = m. a which implies for two bodies respectively

𝑚1 ẍ1 = 𝑘2 (𝑥2 – 𝑥1 ) + 𝐶2 (ẋ2 − ẋ1 )– 𝑘1 𝑥1 − 𝐶1 𝑥1 + 𝐹1 (𝑡),

and

𝑚2 ẍ2 = −𝑘2 (𝑥2 – 𝑥1 ) − 𝐶2 (ẋ2 − ẋ1 ) + 𝐹2 (𝑡).

14
Manipulations yield that;

𝑚1 ẍ1 + (𝐶1 + 𝐶2 )ẋ1 − 𝐶2 ẋ2 + (𝑘1 + 𝑘2 )𝑥1 − 𝑘2 𝑥2 = 𝐹1 (𝑡)

𝑚2 ẍ2 − 𝐶2 ẋ1 + 𝐶2 ẋ2 – 𝑘2 𝑥1 + 𝑘2 𝑥2 = 𝐹2 (𝑡)

This is a system of two second order linear differential equations. There have been many other
physical applications that produce similar systems of linear equations.

There are various solution techniques for such linear systems of differential equations. One
approach is to decouple the system, and obtain a number of differential equations with only one
dependent variable. This approach inevitably increases the order of differential equations. Another
approach is to decrease the order of differential equations to first order which increases the number
of dependent variables, and equations in the set. Some of these techniques will be explained in this
section.

2.1 Decoupling Method

This is probably one of the most popular techniques that decouple the system of coupled equations
into several higher order differential equations in each of the dependent variables. These higher
order linear differential equations are then solved individually.

Example 2.1

Solve the system of the two differential equations

𝑑𝑦
+ z = 𝑒𝑡;
𝑑𝑡
𝑑𝑧
𝑦+ = 𝑒 −𝑡
𝑑𝑡
𝑑𝑦 𝑑𝑦 𝑑2 𝑦 𝑑𝑧
+ z = 𝑒𝑡 → = − z + 𝑒𝑡 → =− + 𝑒𝑡
𝑑𝑡 𝑑𝑡 𝑑𝑡 2 𝑑𝑡
𝑑𝑧 𝑑𝑧 𝑑2 𝑧 𝑑𝑦
𝑦 + 𝑑𝑡 = 𝑒 −𝑡 → = −𝑦 + 𝑒 −𝑡 → =− −𝑒 −𝑡
𝑑𝑡 𝑑𝑡 2 𝑑𝑡

then,

15
 𝑑2𝑦 𝑑2𝑦
= y−𝑒 + 𝑒 → 2 − y = −𝑒 −𝑡 + 𝑒 𝑡
−𝑡 𝑡
𝑑𝑡 2 𝑑𝑡
𝑑2𝑧 𝑡 −𝑡
𝑑2𝑧
= z − 𝑒 −𝑒 → − z = −𝑒 𝑡 −𝑒 −𝑡
𝑑𝑡 2 𝑑𝑡 2

These two 2nd order linear differential equations are to be solved individually. Let’s solve
homogeneous equations first;

𝑦 ′′ − 𝑦 = 0 𝑌ℎ = 𝑒 𝑚𝑡 𝑌ℎ′′ = 𝑚2 𝑒 𝑚𝑡 (𝑚2 − 1)𝑒 𝑚𝑡 = 0

Let
𝑧 ′′ − 𝑧 = 0 𝑍ℎ = 𝑒 𝑚𝑡 𝑍ℎ′′ = 𝑚2 𝑒 𝑚𝑡 𝑚1,2 = ±1

𝑌ℎ = 𝐶1 𝑒 𝑡 +𝐶2 𝑒 −𝑡
𝑍ℎ = 𝐶3 𝑒 𝑡 +𝐶4 𝑒 −𝑡

The general solution of a system of the two first order linear differential equations contains
exactly two arbitrary constants. The relations between the four arbitrary constants in the
homogeneous solution are obtained from homogeneous form of the given set of equations

𝑑𝑦
+ z = 0;
𝑑𝑡
𝑑𝑧
𝑦+ =0
𝑑𝑡

Substitution yields that

𝑦𝑖𝑒𝑙𝑑𝑠
𝐶1 𝑒 𝑡 −𝐶2 𝑒 −𝑡 + 𝐶3 𝑒 𝑡 +𝐶4 𝑒 −𝑡 = 0 → (𝐶1 + 𝐶3 )𝑒 𝑡 + (−𝐶2 +𝐶4 )𝑒 −𝑡 = 0
𝑦𝑖𝑒𝑙𝑑𝑠
𝐶1 𝑒 𝑡 +𝐶2 𝑒 −𝑡 + 𝐶3 𝑒 𝑡 −𝐶4 𝑒 −𝑡 = 0 → (𝐶1 + 𝐶3 )𝑒 𝑡 + (𝐶2 −𝐶4 )𝑒 −𝑡 = 0

𝐶3 = −𝐶1 &𝐶4 = 𝐶2 .

So homogeneous solutions with two arbitrary constants is obtained;

𝑦ℎ = 𝐶1 𝑒 𝑡 +𝐶2 𝑒 −𝑡
𝑧ℎ = −𝐶1 𝑒 𝑡 +𝐶2 𝑒 −𝑡

Particular solution can be proposed based on RHS of the equations as seen in Chapter 1.

𝑦𝑝 = 𝑡(𝐴1 𝑒 𝑡 +𝐴2 𝑒 −𝑡 )

𝑦𝑝′ = 𝐴1 𝑒 𝑡 +𝐴2 𝑒 −𝑡 + 𝑡(𝐴1 𝑒 𝑡 −𝐴2 𝑒 −𝑡 )

𝑦𝑝′′ = 𝐴1 𝑒 𝑡 −𝐴2 𝑒 −𝑡 + 𝐴1 𝑒 𝑡 − 𝐴2 𝑒 −𝑡 + 𝑡(𝐴1 𝑒 𝑡 +𝐴2 𝑒 −𝑡 )

16
 Substitution of these into 𝑦 ′′ − 𝑦 = 𝑒 𝑡 − 𝑒 −𝑡 yields that

(𝐴1 + 𝐴1 + 𝐴1 𝑡 − 𝐴1 𝑡 − 1)𝑒 𝑡 + (−𝐴2 −𝐴2 + 𝐴2 𝑡−𝐴2 𝑡 + 1)𝑒 −𝑡 = 0

1 1
𝐴1 = &𝐴2 =
2 2
1
𝑦𝑝 = 2 𝑡(𝑒 𝑡 + 𝑒 −𝑡 ) Since

1
𝑦𝑔𝑒𝑛 = 𝑦ℎ + 𝑦𝑝 = 𝐶1 𝑒 𝑡 +𝐶2 𝑒 −𝑡 + 𝑡(𝑒 𝑡 + 𝑒 −𝑡 )
2

Similarly, particular solution for z is proposed as

𝑧𝑝 = 𝑡(𝐵1 𝑒 𝑡 +𝐵2 𝑒 −𝑡 )

𝑧𝑝′′ = 𝐵1 𝑒 𝑡 −𝐵2 𝑒 −𝑡 + 𝐵1 𝑒 𝑡 − 𝐵2 𝑒 −𝑡 + 𝑡(𝐵1 𝑒 𝑡 +𝐵2 𝑒 −𝑡 )

Substitution of these into 𝑧 ′′ − 𝑧 = −𝑒 𝑡 − 𝑒 −𝑡 yields that

(𝐵1 + 𝐵1 + 𝐵1 𝑡 − 𝐵1 𝑡 + 1)𝑒 𝑡 + (−𝐵2 −𝐵2 + 𝐵2 𝑡−𝐵2 𝑡 + 1)𝑒 −𝑡 = 0

1 1 1
𝐵1 = − 2 &𝐵2 = 2 then 𝑧𝑝 = 2 𝑡(𝑒 −𝑡 − 𝑒 𝑡 )

1
𝑧𝑔𝑒𝑛 = −𝐶1 𝑒 𝑡 +𝐶2 𝑒 −𝑡 + 𝑡(𝑒 −𝑡 − 𝑒 𝑡 )
2

Another strategy to find 𝑧𝑔𝑒𝑛 is to use 𝑦𝑔𝑒𝑛 in

𝑑𝑦𝑔𝑒𝑛
+ 𝑧𝑔𝑒𝑛 = 𝑒 𝑡 .
𝑑𝑡

𝑑𝑦𝑔𝑒𝑛 1
𝑧𝑔𝑒𝑛 = 𝑒 𝑡 − = −𝐶1 𝑒 𝑡 +𝐶2 𝑒 −𝑡 + 𝑡(𝑒 −𝑡 − 𝑒 𝑡 ).
𝑑𝑡 2

Example 2.2

Solve the system of the two first order linear differential equations

17
 𝑑𝑥
+ y=2
𝑑𝑡
𝑑𝑦
𝑥 + 𝑑𝑡 = cos 𝑡

𝑑𝑥
+ y = 2, 𝑥′ = − y + 2 𝑥 ′′ = −𝑦 ′ , 𝑥 ′′ − 𝑥 = − cos 𝑡
𝑑𝑡

𝑑𝑦
𝑥+ = cos 𝑡, y ′ = −x + cos 𝑡 𝑦 ′′ = −𝑥 ′ − sin 𝑡 , 𝑦 ′′ − 𝑦 = −2 − sin 𝑡
𝑑𝑡

Let’s solve them individually.

Let’s start with the two corresponding homogeneous equations;

𝑥 ′′ − 𝑥 = 0 𝑥ℎ = 𝑒 𝑚𝑡 ⟹ 𝑥ℎ′′ = 𝑚2 𝑒 𝑚𝑡 ⟹ 𝑚1,2 = ±1
Let
𝑦 ′′ − 𝑦 = 0 𝑦ℎ = 𝑒 𝑚𝑡 ⟹ 𝑦ℎ′′ = 𝑚2 𝑒 𝑚𝑡 ⟹ 𝑚1,2 = ±1

𝑥ℎ = 𝐶1 𝑒 𝑡 +𝐶2 𝑒 −𝑡
𝑦ℎ = 𝐶3 𝑒 𝑡 +𝐶4 𝑒 −𝑡

For the first second order equation

𝑥 ′′ − 𝑥 = − cos 𝑡

since the right side is {-cost} let a particular solution be;

𝑥𝑝 = 𝐴 cos 𝑡 + 𝐵 sin 𝑡

Then

𝑥𝑝′ = −𝐴 sin 𝑡 + 𝐵 cos 𝑡

𝑥𝑝′′ = −𝐴 cos 𝑡 − 𝐵 sin 𝑡

Substitution of these into 𝑥 ′′ − 𝑥 = − cos 𝑡 yields that

−𝐴 cos 𝑡 − 𝐵 sin 𝑡 − 𝐴 cos 𝑡 − 𝐵 sin 𝑡 = − cos 𝑡

𝑦𝑖𝑒𝑙𝑑𝑠 𝐴 = 1⁄2 𝑦𝑖𝑒𝑙𝑑𝑠 1

−2𝐴 cos 𝑡 − 2𝐵 sin 𝑡 = − cos 𝑡 → → 𝑥𝑝 = cos 𝑡
𝐵=0 2

Hence

18
 1
𝑥𝑔𝑒𝑛 = 𝑥ℎ + 𝑥𝑝 = 𝐶1 𝑒 𝑡 +𝐶2 𝑒 −𝑡 + cos 𝑡.
2

Consider the other equation; 𝑦 ′′ − 𝑦 = − sin 𝑡 − 2,

let a particular solution be;

𝑦𝑝 = 𝐶 cos 𝑡 + 𝐷 sin 𝑡 + 𝐸.

Then

𝑦𝑝′ = −𝐶 sin 𝑡 + 𝐷 cos 𝑡

𝑦𝑝′′ = −𝐶 cos 𝑡 − 𝐷 sin 𝑡

Substitution of these into 𝑦 ′′ − 𝑦 = − sin 𝑡 − 2 yields that

−𝐶 cos 𝑡 − 𝐷 sin 𝑡 − 𝐶𝑐𝑜𝑠𝑡 − 𝐷 sin 𝑡 − 𝐸 + sin 𝑡 + 2 = 0

−2𝐶 cos 𝑡 + (−2𝐷 + 1) sin 𝑡 + (−𝐸 + 2) = 0

1 𝑦𝑖𝑒𝑙𝑑𝑠 1
𝐶 = 0, 𝐷= & 𝐸=2 → 𝑦𝑝 = sin 𝑡 + 2
2 2

Hence

1
𝑦𝑔𝑒𝑛 = 𝑦ℎ + 𝑦𝑝 = 𝐶3 𝑒 𝑡 +𝐶4 𝑒 −𝑡 + sin 𝑡 + 2
2

The general solution of a linear system of two first order differential equations contains exactly
two arbitrary constants. Now we will substitute the two solutions into one of the equations

𝑑𝑦
𝑥+ = cos 𝑡
𝑑𝑡

in the given system to eliminate extra arbitrary constants.

Recall
1
𝑥𝑔𝑒𝑛 = 𝐶1 𝑒 𝑡 +𝐶2 𝑒 −𝑡 + 2 cos 𝑡

𝑑𝑦𝑔𝑒𝑛 1
= 𝐶3 𝑒 𝑡 −𝐶4 𝑒 −𝑡 + 2 cos 𝑡
𝑑𝑡

Substitution yields that

19
 1 1
𝐶1 𝑒 𝑡 +𝐶2 𝑒 −𝑡 + cos 𝑡 + 𝐶3 𝑒 𝑡 −𝐶4 𝑒 −𝑡 + cos 𝑡 − cos 𝑡 = 0
2 2
1 1 𝑦𝑖𝑒𝑙𝑑𝑠 𝐶3 = −𝐶1
(𝐶1 + 𝐶3 )𝑒 𝑡 +(𝐶2 −𝐶4 )𝑒 −𝑡 + ( + − 1) 𝑐𝑜𝑠𝑡 = 0 →
2 2 𝐶4 = 𝐶2

Therefore general solution of the two unknowns are:

1
𝑥𝑔𝑒𝑛 = 𝐶1 𝑒 𝑡 +𝐶2 𝑒 −𝑡 + cos 𝑡
2
1
𝑦𝑔𝑒𝑛 = −𝐶1 𝑒 𝑡 +𝐶2 𝑒 −𝑡 + sin 𝑡 + 2
2
𝑑𝑦
This result may be verified using the other differential equation + 𝑥 = 2 . The same
𝑑𝑡

result will be obtained.

After finding the solution 𝑥𝑔𝑒𝑛 , a simpler way to find 𝑦𝑔𝑒𝑛 is to substitute 𝑥𝑔𝑒𝑛 in the first
𝑑𝑥
equation + y = 2 which leads
𝑑𝑡

𝑑𝑥𝑔𝑒𝑛 1
𝑦𝑔𝑒𝑛 = 2 − = −𝐶1 𝑒 𝑡 +𝐶2 𝑒 −𝑡 + sin 𝑡 + 2.
𝑑𝑡 2

Example 2.3

Solve the system of the two first order linear differential equations

𝑑𝑥 1
−y=
𝑑𝑡 sin 𝑡
𝑑𝑦 1
𝑥+ =
𝑑𝑡 cos 𝑡
1 cos 𝑡 1 cos 𝑡
𝑥′ = 𝑦 + , 𝑥"= 𝑦 ′ − 2 , 𝑥" = −𝑥 + − 2
sin 𝑡 sin t cos 𝑡 sin t

1 sin 𝑡 1 sin 𝑡
𝑦 ′ = −𝑥 + , 𝑦"= − 𝑥 ′ + 2
, 𝑦" = −𝑦 ′ − +
cos 𝑡 𝑐𝑜𝑠 𝑡 sin 𝑡 𝑐𝑜𝑠 2 𝑡

The two uncoupled equations are

20
 1 cos 𝑡
𝑥" + 𝑥 = − 2
cos 𝑡 sin t

and

1 sin 𝑡
𝑦" + 𝑦 ′ = − + .
sin 𝑡 𝑐𝑜𝑠 2 𝑡

Homogeneous solutions;

𝑥 ′′ + 𝑥 = 0 𝑥ℎ = 𝑒 𝑚𝑡 ⟹ 𝑥ℎ′′ = 𝑚2 𝑒 𝑚𝑡 ⟹ 𝑚1,2 = ±𝑖
Let
𝑦 ′′ + 𝑦 = 0 𝑦ℎ = 𝑒 𝑚𝑡 ⟹ 𝑦ℎ′′ = 𝑚2 𝑒 𝑚𝑡 ⟹ 𝑚1,2 = ±𝑖

𝑥ℎ = 𝐶1 cos 𝑡 + 𝐶2 sin 𝑡
𝑦ℎ = 𝐶3 cos 𝑡 + 𝐶4 sin 𝑡

These should satisfy homogeneous equations. Therefore one has

𝐶4 = −𝐶1
𝐶3 = 𝐶2

2.2 Operator Method:

This is probably one of the most popular techniques that eliminate the number of dependent
parameters to a single variable of higher order differential equations. In operator method each
differentiation is denoted by a linear operator, D, and the resulting algebraic equation set is solved
in terms of D using linear algebra rules. The unknowns are solved in terms of D. Since D is a
differential operator, higher order independent linear differential equations are then obtained and
solved individually.

Example
𝑑𝑦
+ z = 𝑒𝑡
𝑑𝑡
Solve the differential equation system 𝑑𝑧
𝑦 + 𝑑𝑡 = 𝑒 −𝑡

𝑑
Let 𝐷= then,
𝑑𝑡

21
 𝐷𝑦 + z = 𝑒 𝑡 𝐷 1
⟹ ∆= | | = D2 − 1
𝑦 + 𝐷𝑧 = 𝑒 −𝑡 1 𝐷
𝑡
∆𝑦 = | 𝑒−𝑡 1 | = 𝐷𝑒 𝑡 − 𝑒 −𝑡 = 𝑒 𝑡 − 𝑒 𝑡
𝑒 𝐷

∆𝑧 = |𝐷 𝑒 𝑡 | = 𝐷𝑒 −𝑡 − 𝑒 −𝑡 = −𝑒 −𝑡 − 𝑒 𝑡
1 𝑒 −𝑡

∆𝑦 𝑒 𝑡 − 𝑒 𝑡 𝑦𝑖𝑒𝑙𝑑𝑠 𝑦𝑖𝑒𝑙𝑑𝑠
𝑦= = 2 → (𝐷 2 − 1)𝑦 = 𝑒 𝑡 − 𝑒 −𝑡 → 𝑦 ′′ − 𝑦 = 𝑒 𝑡 − 𝑒 −𝑡
∆ 𝐷 −1

∆𝑧 −𝑒 −𝑡 − 𝑒 𝑡 𝑦𝑖𝑒𝑙𝑑𝑠 𝑦𝑖𝑒𝑙𝑑𝑠
𝑧= = 2
→ (𝐷 2 − 1)𝑧 = −𝑒 −𝑡 − 𝑒 𝑡 → 𝑧 ′′ − 𝑧 = −𝑒 𝑡 − 𝑒 −𝑡
∆ 𝐷 −1

These two 2nd order differential equations are to be solved individually. Let’s solve homogeneous
equations first;

𝑦 ′′ − 𝑦 = 0 𝑌ℎ = 𝑒 𝑚𝑡 𝑌ℎ′′ = 𝑚2 𝑒 𝑚𝑡 (𝑚2 − 1)𝑒 𝑚𝑡 = 0

Let
𝑧 ′′ − 𝑧 = 0 𝑍ℎ = 𝑒 𝑚𝑡 𝑍ℎ′′ = 𝑚2 𝑒 𝑚𝑡 𝑚1,2 = ±1

𝑌ℎ = 𝐶1 𝑒 𝑡 +𝐶2 𝑒 −𝑡
𝑍ℎ = 𝐶3 𝑒 𝑡 +𝐶4 𝑒 −𝑡

The relation between coefficients are obtained from homogeneous form of the given equation set

𝐷𝑦ℎ = 𝐶1 𝑒 𝑡 −𝐶2 𝑒 −𝑡
Substitution yields that
𝐷𝑧ℎ = 𝐶3 𝑒 𝑡 −𝐶4 𝑒 −𝑡
𝑦𝑖𝑒𝑙𝑑𝑠 𝑦𝑖𝑒𝑙𝑑𝑠
𝐷𝑦 + 𝑧 = 𝑒 𝑡 → 𝐶1 𝑒 𝑡 −𝐶2 𝑒 −𝑡 + 𝐶3 𝑒 𝑡 +𝐶4 𝑒 −𝑡 = 0 → (𝐶1 + 𝐶3 )𝑒 𝑡 + (−𝐶2 +𝐶4 )𝑒 −𝑡 = 0
𝑦𝑖𝑒𝑙𝑑𝑠 𝑦𝑖𝑒𝑙𝑑𝑠
𝑦 + 𝐷𝑧 = 𝑒 −𝑡 → 𝐶1 𝑒 𝑡 +𝐶2 𝑒 −𝑡 + 𝐶3 𝑒 𝑡 −𝐶4 𝑒 −𝑡 = 0 → (𝐶1 + 𝐶3 )𝑒 𝑡 + (𝐶2 −𝐶4 )𝑒 −𝑡 = 0

𝐶3 = −𝐶1 & 𝐶4 = 𝐶3 .

So homogeneous solutions;

𝑌ℎ = 𝐶1 𝑒 𝑡 +𝐶2 𝑒 −𝑡
𝑍ℎ = −𝐶1 𝑒 𝑡 +𝐶2 𝑒 −𝑡

Particular solution can be proposed based on RHS of the equation;

𝑌𝑝 = 𝑡(𝐴1 𝑒 𝑡 +𝐴2 𝑒 −𝑡 )

𝑌𝑝′ = 𝐴1 𝑒 𝑡 +𝐴2 𝑒 −𝑡 + 𝑡(𝐴1 𝑒 𝑡 −𝐴2 𝑒 −𝑡 )

22
 𝑌𝑝′′ = 𝐴1 𝑒 𝑡 −𝐴2 𝑒 −𝑡 + 𝐴1 𝑒 𝑡 − 𝐴2 𝑒 −𝑡 + 𝑡(𝐴1 𝑒 𝑡 +𝐴2 𝑒 −𝑡 )

Substitution of these into 𝑦 ′′ − 𝑦 = 𝑒 𝑡 − 𝑒 −𝑡 yields that

(𝐴1 + 𝐴1 + 𝐴1 𝑡 − 𝐴1 𝑡 − 1)𝑒 𝑡 + (−𝐴2 −𝐴2 + 𝐴2 𝑡−𝐴2 𝑡 + 1)𝑒 −𝑡 = 0

1 1
𝐴1 = & 𝐴2 =
2 2
1
𝑌𝑝 = 2 𝑡(𝑒 𝑡 + 𝑒 −𝑡 ) Since

1
𝑌𝑔𝑒𝑛 = 𝑌ℎ + 𝑌𝑝 = 𝐶1 𝑒 𝑡 +𝐶2 𝑒 −𝑡 + 𝑡(𝑒 𝑡 + 𝑒 −𝑡 )
2

Similarly, particular solution for Z is proposed to be

𝑍𝑝 = 𝑡(𝐵1 𝑒 𝑡 +𝐵2 𝑒 −𝑡 )

𝑍𝑝′′ = 𝐵1 𝑒 𝑡 −𝐵2 𝑒 −𝑡 + 𝐵1 𝑒 𝑡 − 𝐵2 𝑒 −𝑡 + 𝑡(𝐵1 𝑒 𝑡 +𝐵2 𝑒 −𝑡 )

Substitution of these into 𝑧 ′′ − 𝑧 = −𝑒 𝑡 − 𝑒 −𝑡 yields that

(𝐵1 + 𝐵1 + 𝐵1 𝑡 − 𝐵1 𝑡 + 1)𝑒 𝑡 + (−𝐵2 −𝐵2 + 𝐵2 𝑡−𝐵2 𝑡 + 1)𝑒 −𝑡 = 0

1 1 1
𝐵1 = − 2 & 𝐵2 = 2 then 𝑍𝑝 = 2 𝑡(𝑒 −𝑡 − 𝑒 𝑡 )

1
𝑍𝑔𝑒𝑛 = −𝐶1 𝑒 𝑡 +𝐶2 𝑒 −𝑡 + 𝑡(𝑒 −𝑡 − 𝑒 𝑡 )
2

Example
𝑑𝑥
+ y=2
𝑑𝑡
Solve the given differential equations 𝑑𝑦
𝑥 + 𝑑𝑡 = cos 𝑡

𝑑
𝐷= , substituting D into the system;
𝑑𝑡

𝐷𝑥 + 𝑦 = 2 𝐷 1
⟹ ∆= | | = 𝐷2 − 1
𝑥 + 𝐷𝑦 = cos 𝑡 1 𝐷

2 1
∆𝑥 = | | = 𝐷2 − cos 𝑡 = − cos 𝑡
𝑐𝑜𝑠𝑡 𝐷

23
 𝐷 2
∆𝑦 = | | = 𝐷 cos 𝑡 − 2 = − sin 𝑡 − 2
1 cos 𝑡
∆𝑥 − cos 𝑡 𝑦𝑖𝑒𝑙𝑑𝑠 𝑦𝑖𝑒𝑙𝑑𝑠
x= = 2 → (D2 − 1)x = − cos 𝑡 → 𝑥 ′′ − 𝑥 = − cos 𝑡
∆ D −1

∆𝑦 − sin 𝑡 − 2 𝑦𝑖𝑒𝑙𝑑𝑠 𝑦𝑖𝑒𝑙𝑑𝑠

y= = → (D2 − 1)y = − sin 𝑡 − 2 → 𝑦 ′′ − 𝑦 = − sin 𝑡 − 2
∆ D2 − 1

Let’s solve them individually. Let’s start with the corresponding homogeneous system;

𝑥 ′′ − 𝑥 = 0 𝑥ℎ = 𝑒 𝑚𝑡 ⟹ 𝑥ℎ′′ = 𝑚2 𝑒 𝑚𝑡 ⟹ 𝑚2 − 1 =0
Let
𝑦 ′′ − 𝑦 = 0 𝑦ℎ = 𝑒 𝑚𝑡 ⟹ 𝑦ℎ′′ = 𝑚2 𝑒 𝑚𝑡 ⟹ 𝑚1,2 = ±1

𝑥ℎ = 𝐶1 𝑒 𝑡 +𝐶2 𝑒 −𝑡
𝑦ℎ = 𝐶3 𝑒 𝑡 +𝐶4 𝑒 −𝑡

Since the right side is {-cost} let a particular solution be;

𝑥𝑝 = 𝐴 cos 𝑡 + 𝐵 sin 𝑡

𝑥𝑝′ = −𝐴 sin 𝑡 + 𝐵 cos 𝑡

𝑥𝑝′′ = −𝐴 cos 𝑡 − 𝐵 sin 𝑡

Substitution of these into 𝑥 ′′ − 𝑥 = − cos 𝑡 yields that

−𝐴 cos 𝑡 − 𝐵 sin 𝑡 − 𝐴 cos 𝑡 + 𝐵 sin 𝑡 = − cos 𝑡

𝑦𝑖𝑒𝑙𝑑𝑠 𝐴 = 1⁄2 𝑦𝑖𝑒𝑙𝑑𝑠 1

−2𝐴 cos 𝑡 − 2𝐵 sin 𝑡 = − cos 𝑡 → → 𝑥𝑝 = cos 𝑡
𝐵=0 2
1
𝑥𝑔𝑒𝑛 = 𝑥ℎ + 𝑥𝑝 = 𝐶1 𝑒 𝑡 +𝐶2 𝑒 −𝑡 + cos 𝑡
2

Consider the other equation; 𝑦 ′′ − 𝑦 = − sin 𝑡 − 2, let a particular solution be;

𝑦𝑝 = 𝐶 cos 𝑡 + 𝐷 sin 𝑡 + 𝐸

𝑦𝑝′ = −𝐶 sin 𝑡 + 𝐷 cos 𝑡

𝑦𝑝′′ = −𝐶 cos 𝑡 − 𝐷 sin 𝑡

Substitution of these into 𝑦 ′′ − 𝑦 = − sin 𝑡 − 2 yields that

−𝐶 cos 𝑡 − 𝐷 sin 𝑡 − 𝐶𝑐𝑜𝑠𝑡 + 𝐷 sin 𝑡 + 𝐸 + sin 𝑡 + 2 = 0

24
 −2𝐶 cos 𝑡 + (−2𝐷 + 1) sin 𝑡 + (−𝐸 + 2) = 0

1 𝑦𝑖𝑒𝑙𝑑𝑠 1
𝐶 = 0, 𝐷= & 𝐸=2 → 𝑦𝑝 = sin 𝑡 + 2
2 2
1
𝑦𝑔𝑒𝑛 = 𝑦ℎ + 𝑦𝑝 = 𝐶3 𝑒 𝑡 +𝐶4 𝑒 −𝑡 + sin 𝑡 + 2
2

Now we should substitute the solution into the given set to eliminate the dependent parameters.
1
𝑥 = 𝐶1 𝑒 𝑡 +𝐶2 𝑒 −𝑡 + 2 cos 𝑡
𝑑𝑦
𝑥 + 𝑑𝑡 = cos 𝑡 where
𝑑𝑦 1
= 𝐶3 𝑒 𝑡 −𝐶4 𝑒 −𝑡 + 2 cos 𝑡
𝑑𝑡

Substitution yields that

1 1
𝐶1 𝑒 𝑡 +𝐶2 𝑒 −𝑡 + cos 𝑡 + 𝐶3 𝑒 𝑡 −𝐶4 𝑒 −𝑡 + cos 𝑡 − cos 𝑡 = 0
2 2
1 1 𝑦𝑖𝑒𝑙𝑑𝑠 𝐶3 = −𝐶1
(𝐶1 + 𝐶3 )𝑒 𝑡 +(𝐶2 −𝐶4 )𝑒 −𝑡 + ( + − 1) 𝑐𝑜𝑠𝑡 = 0 →
2 2 𝐶4 = 𝐶2

1
𝑥𝑔𝑒𝑛 = 𝐶1 𝑒 𝑡 +𝐶2 𝑒 −𝑡 + cos 𝑡
2
1
𝑦𝑔𝑒𝑛 = −𝐶1 𝑒 𝑡 +𝐶2 𝑒 −𝑡 + sin 𝑡 + 2
2
𝑑𝑦
This result may be verified using the other differential equation + 𝑥 = 2 . The same
𝑑𝑡

result will be obtained.

Example
𝑑𝑥 1
− y = sin 𝑡
𝑑𝑡
Solve the differential equations 𝑑𝑦 1
𝑥 + 𝑑𝑡 = cos 𝑡

1
𝐷𝑥 − y =
sin 𝑡 ⟹ ∆= |𝐷 −1
| = 𝐷2 + 1
1 1 𝐷
𝑥 + 𝐷𝑦 =
cos 𝑡

25
 1
−1 1 1 cos 𝑡 1
∆𝑥 = | sin 𝑡 |=𝐷 + =− +
1 sin 𝑡 cos 𝑡 𝑠𝑖𝑛2 𝑡 cos 𝑡
𝐷
cos 𝑡
1
D
∆𝑦 = | sin 𝑡 | = 𝐷 1 + 1 = − sin 𝑡 − 1
1 cos 𝑡 sin 𝑡 𝑐𝑜𝑠 2 𝑡 sin 𝑡
1
cos 𝑡
∆𝑥 ∆𝑥 𝑦𝑖𝑒𝑙𝑑𝑠 𝑦𝑖𝑒𝑙𝑑𝑠 cos 𝑡 1
x= = 2 → (D2 + 1)x = ∆𝑥 → 𝑥 ′′ + 𝑥 = − 2
+
∆ D +1 𝑠𝑖𝑛 𝑡 cos 𝑡
∆𝑦 ∆𝑦 𝑦𝑖𝑒𝑙𝑑𝑠 𝑦𝑖𝑒𝑙𝑑𝑠 sin 𝑡 1
y= = 2 → (D2 + 1)y = ∆𝑦 → 𝑦 ′′ + 𝑦 = − 2
−
∆ D +1 𝑐𝑜𝑠 𝑡 sin 𝑡

Homogeneous solutions;

𝑥 ′′ + 𝑥 = 0 𝑥ℎ = 𝑒 𝑚𝑡 ⟹ 𝑥ℎ′′ = 𝑚2 𝑒 𝑚𝑡 ⟹ 𝑚2 + 1 =0
Let
𝑦 ′′ + 𝑦 = 0 𝑦ℎ = 𝑒 𝑚𝑡 ⟹ 𝑦ℎ′′ = 𝑚2 𝑒 𝑚𝑡 ⟹ 𝑚1,2 = ±𝑖

𝑥ℎ = 𝐶1 cos 𝑡 + 𝐶2 sin 𝑡
𝑦ℎ = 𝐶3 cos 𝑡 + 𝐶4 sin 𝑡

Example 2.4 Consider two masses connected to the walls by springs, and they are connected to
each other by massless springs. Suppose that 𝑚1 = 𝑚2 = 𝑘1 = 𝑘2 = 𝑘3 = 1. Find 𝑥1 (𝑡) and
𝑥2 (𝑡) are the deviation from equilibrium for both masses.

Figure.2.2Spring and mass system.

Newton’s Second Law of Motion yields

26
Example 2.5

Solve the differential equations

𝑥̇ − 𝑥 − 𝑦 = 3𝑡
𝑥̇ + 𝑦̇ − 5𝑥 − 2𝑦 = 5

𝑥̇ − 𝑥 = 𝑦 + 3𝑡 (1)

𝑥̇ − 5𝑥 = 2𝑦 − 𝑦̇ + 5 (2)

From (1) solve y, compute 𝑦̇

y = 𝑥̇ − 𝑥 − 3𝑡 → 𝑦̇ = ẍ − 𝑥̇ − 3

Substitute in (2) get the decoupled differential equation for x

𝑥̇ − 5𝑥 = 2𝑦 − 𝑦̇ + 5 = 2(𝑥̇ − 𝑥 − 3𝑡) − (ẍ − 𝑥̇ − 3) + 5

→ ẍ − 2𝑥̇ − 3x = 8 − 6t

From these equations (1) and (2) we solve 𝑥, 𝑥̇

4𝑥 = 𝑦̇ − y + 3t − 5 → 4𝑥̇ = ÿ − 𝑦̇ + 3

4𝑥̇ = 3𝑦 + 𝑦̇ + 15𝑡 − 5

Equating the two equations, we get the decoupled differential equation for y

ÿ − 𝑦̇ + 3 = 𝑦̇ + 3y − 5 + 15t

→ ÿ − 2𝑦̇ − 3y = −8 + 15t

Solution of x-equation

ẍ − 2𝑥̇ − 3x = 8 − 6t

Homogeneous solution;

𝑟1 = −1
𝑟 2 − 2𝑟 − 3 = 0 ⟹ ⟹ 𝑥ℎ = 𝑐1 𝑒 −𝑡 + 𝑐2 𝑒 3𝑡
𝑟2 = 3

Particular solution;

𝑥𝑝 = A + Bt → 𝑥̇ 𝑝 = B → ẍ 𝑝 = 0

ẍ − 2𝑥̇ − 3x = 8 − 6t → −2𝐵 − 3(A + Bt) = 8 − 6t

27
 −3𝐵 = −6 → B = 2

−2𝐵 − 3𝐴 = 8 → A = −4

𝑥𝑝 = 2𝑡 − 4

𝑥𝑔𝑒𝑛 = 2𝑡 − 4 + 𝑐1 𝑒 −𝑡 + 𝑐2 𝑒 3𝑡

Solution of y-equation

ÿ − 2𝑦̇ − 3y = −8 + 15t

Homogeneous solution;

𝑟1 = −1
𝑟 2 − 2𝑟 − 3 = 0 ⟹ ⟹ 𝑦ℎ = 𝑐3 𝑒 −𝑡 + 𝑐4 𝑒 3𝑡
𝑟2 = 3

Particular solution;

𝑦𝑝 = A + Bt → 𝑦̇𝑝 = B → ÿ 𝑝 = 0

ÿ − 2𝑦̇ − 3y = −8 + 15t → −2𝐵 − 3(A + Bt) = −8 + 15t

−3𝐵 = 15 → B = −5

−2𝐵 − 3𝐴 = −8 → A = 6

𝑦𝑝 = 6 − 5t

𝑦𝑔𝑒𝑛 = 6 − 5t + 𝑐3 𝑒 −𝑡 + 𝑐4 𝑒 3𝑡

To remove extra arbitrary constants we use one of the equations in the system

𝑥̇ − 𝑥 = 𝑦 + 3𝑡

which leads

2 − 𝑒 −𝑡 𝑐1 + 3𝑒 3𝑡 𝑐2 − (2𝑡 − 4 + 𝑐1 𝑒 −𝑡 + 𝑐2 𝑒 3𝑡 ) = 6 − 5t + 𝑐3 𝑒 −𝑡 + 𝑐4 𝑒 3𝑡 + 3t

Therefore 𝐶3 = −2𝐶1 , 𝐶4 = 2𝐶2 , and hence

𝑦𝑔𝑒𝑛 = 6 − 5t − 2𝑐1 𝑒 −𝑡 − 2𝑐2 𝑒 3𝑡

28
 ∎

Homeworks
2𝑒 −𝑡 −3𝑡
𝑥 ′ + 3𝑥 =
√2
1.
′ 2𝑒 −𝑡 +3𝑡
y +𝑦 =
√2
′
𝑥 − 2𝑥 − 𝑦 = −1, 𝑥 (0) = 1
2.
y ′ + 𝑥 − 2𝑦 = 8, 𝑦 (0) = 1
𝑥 ′ − 3𝑥 = −4 sin 2𝑡 , 𝑥(0) = 2
3.
y ′ − 5𝑥 + 2𝑦 = cos 2𝑡 , 𝑦(0) = −1
𝑥̇ − 𝑥 − 2𝑦 = 2𝑡
4.
𝑦̇ − 3𝑥 − 2𝑦 = −4t
𝑥 ′ + 5𝑥 − 𝑦 = 6𝑒 2𝑡
5.
y ′ − 4𝑥 + 2𝑦 = −𝑒 2𝑡 − 4𝑡

29