System of First Order Equations

Anil Kumar
A-408

System of 1st order ODE

} General form of the system of simultaneous first order
ordinary differential equations
x1¢ = f1 (t , x1 , x2 , xn )
x2¢ = f 2 (t , x1 , x2 , xn )

xn¢ = f n (t , x1 , x2 , xn )
where each xk is a function of t.
} If each fk is a linear function of x1, x2, …, xn, then the system
of equations is said to be linear, otherwise it is nonlinear.
} Systems of higher order differential equations can similarly
be defined.
•2 •Anil Kumar, Dept. of Mathematics
 Example 1. System of 1st order ODE
} Consider the differential equation
u¢¢(t ) + 16 u¢(t ) + 192u (t ) = 0
} This second order equation can be converted into a system
of first order equations by letting
x1 = u and x2 = u'.
Thus
x1¢ = x2
x2¢ + 16 x2 + 192 x1 = 0

or x1¢ = x2
x2¢ = -16 x2 - 192 x1

•3 •Anil Kumar, Dept. of Mathematics

nth Order ODEs and Linear 1st Order Systems

} The method illustrated in previous example can be used to
transform an arbitrary nth order equation

y( n) = f ( t , y, y¢, y¢¢,, y ( n-1) ) (1)

into a system of n first order equations, first by defining
x1 = y, x2 = y¢, x3 = y¢¢,  , xn = y ( n -1) (2)
Then
x1¢ = x2
x2¢ = x3
 (3)
xn¢ -1 = xn
xn¢ = f (t , x1 , x2 , xn )
•4 •Anil Kumar, Dept. of Mathematics
System of 1st order ODE

} (1) and (3) are equivalent in the sense

if y(t) is a solution of (1), then the functions x1(t), x2(t),

…, xn(t) defined by (2) satisfy (3), and conversely, if

x1(t), x2(t), …, xn(t) satisfy (3), then y(t) is a solution of

(1).

•5 •Anil Kumar, Dept. of Mathematics

Solutions of First Order Systems

} A system of simultaneous first order ordinary differential
equations has the general form
x1¢ = f1 (t , x1 , x2 , xn )
x2' = f 2 (t , x1 , x2 , xn )
 (1)
xn¢ = f n (t , x1 , x2 , xn ).
It has a solution on I: a ≤ t ≤ b if there exists n functions
x1 (t ), x2 (t ),, xn (t )
which are differentiable on I and satisfy the system of
equations at all points t in I.
} Initial conditions may also be prescribed to give an IVP:
x1 (t0 ) = a1 , x2 (t0 ) = a2 ,, xn (t0 ) = an (2)
•6 •Anil Kumar, Dept. of Mathematics
Theorem 1:
(Existence and uniqueness of the solution)

Suppose f1,…, fn and the partial derivatives

¶f1 ¶f ¶f ¶f
,..., 1 ,..., n ,..., n
¶x1 ¶xn ¶x1 ¶xn
are continuous in a region R of (t, x1, x2, …,xn) space. If
(t0 , a1, a2 ,, an ) is an interior point of R. Then the system (1) has
a unique solution x1 (t ), x2 (t ),, xn (t ) that satisfies the initial
conditions (2).

•7 •Anil Kumar, Dept. of Mathematics

System of 1st order ODE

} Theorem 1 is a special case of the following theorem for the

equation (n)
y = f ( t , y, y¢, y¢¢,, y
( n -1)
) (*)

} Theorem 2: Let the function f and the partial derivatives

¶f/¶y, ¶f/¶y',…, ¶f/¶y(n-1) be continuous in a region R of (t, y,
y',…, y(n-1)) space. If ( t0 , a1, a2 ,, an ) is an interior point of R,
then the equation (*) has a unique solution y(t) that satisfies
the initial conditions y(t0) = a1, y'(t0) = a2, …, y(n-1)(t0) = an.

•8 •Anil Kumar, Dept. of Mathematics

 Examples
} Transform the following equations into a system of first
order equations.
1
(i ) y¢¢ + y¢ + 2 y = 0,
2
1
(ii ) y¢¢ + y¢ + 2 y = 3sin t ,
2
æ 1ö
(iii ) t 2 y¢¢ + ty ¢ + ç t 2 - ÷ y = 0,
è 4ø
(iv) y¢¢¢ = y¢¢ - x 2 ( y ¢) 2 ,
(v ) y¢¢ + p(t ) y¢ + q(t ) y = g (t ), y (0) = a 0 and y ¢(0) = a1.

•9 •Anil Kumar, Dept. of Mathematics

Exercises

1. Transform the given system into a single equation of second

order

(i ) x1¢ = x2 , x2¢ = 2 x1 - 2 x2 .
(ii ) x1¢ = x2 , x2¢ = -2 x1.

•10 •Anil Kumar, Dept. of Mathematics

 Linear systems
} The general linear system of first-order differential
equations is of the form

ì dx
ïï dt = a1 (t ) x + b1 (t ) y + f1 (t )
í
ï dy = a (t ) x + b (t ) y + f (t )
ïî dt 2 2 2

where a1(t), b1(t), a2(t), b2(t), f1(t), f2(t) are continuous on a

closed interval I = [a, b] of the t-axis. If f1(t) = f2(t) = 0,
then the system is called homogeneous, otherwise non-
homogeneous.
•Anil Kumar, Dept. of Mathematics •11

Solutions of the linear systems

} A solution of the above system on I is a pair of functions x(t),
y(t) satisfying both the equations throughout the interval.

ì x = x(t )
We shall write such a solution in the form í .
î y = y (t )

ì x = e 4t ì x = e-2t
Example: í and í -2 t
are the solutions of the
î y = e 4t
î y = - e
ì dx
ïï dt = x + 3 y
homogeneous system (on any interval I) í .
dy
ï = 3x + y
ïî dt

•Anil Kumar, Dept. of Mathematics •12

 Example
} The equation
y ¢¢ + y = 0, 0 £ t £ 2p

can be written as system of first order equations by letting y1 = y

and y2 = y'. Thus '
y1 = y2
y2' = - y1
} A solution to this system is
y1 = sin t , y2 = cos t (0 £ t £ 2p )
which is a parametric description for the unit
circle.

•13 •Anil Kumar, Dept. of Mathematics

Linear Systems
Existence and uniqueness theorem:

Theorem 1: If t0 is any point of the interval I and if x0 and y0 are

any real numbers, then the system
ì dx
ïï dt = a1 (t ) x + b1 (t ) y + f1 (t )
í
ï dy = a (t ) x + b (t ) y + f (t )
ïî dt 2 2 2

ì x = x(t )
has one and only one solution í valid throughout I such
that x(t0) = x0 and y(t0) = y0. î y = y (t )

•Anil Kumar, Dept. of Mathematics •14

 Linear Systems
} Task is to study the structure of the solutions of the
homogeneous system

ì dx
ïï dt = a1 (t ) x + b1 (t ) y
í (1)
ï dy = a (t ) x + b (t ) y
ïî dt 2 2

It is obvious that (x = 0, y = 0) is the trivial solution.

Hence, we seek a nontrivial solution.

•15 •Anil Kumar, Dept. of Mathematics

Linear Systems
} Theorem 2: If the above homogeneous system (1) has two
solutions
ì x = x1 (t ) ì x = x2 (t )
í and í (2)
î y = y1 (t ) î y = y2 (t )
ì x = Ax1 (t ) + Bx2 (t )
on I = [a, b], then í (3)
î y = Ay 1 (t ) + By2 (t )
is also a solution of (1) on I for any arbitrary constants A and B.
(Or any linear combination of two solutions of the homogeneous
system (1) is also a solution.
} Can we say (3) contains all the solutions of (1), or is it a general
solution of (1)?
} By theorem (3), it will be the general solution if the constants A
and B can be chosen so as to satisfy ICs.
•16 •Anil Kumar, Dept. of Mathematics
 Linear Systems

} Definition: Wronskian of the two solutions (2) of the homo-

geneous system (1) is defined as

x1 (t ) x 2 (t )
W(t) =
y1 (t ) y 2 (t )

} Theorem 3: If the Wronskian of the two solutions (2) of the

homogeneous system (1) doesn’t vanish on I, then (3) is the
general solution of the given homogeneous system on this
interval.

•17 •Anil Kumar, Dept. of Mathematics

Linear Systems

} Remark: Vanishing or non-vanishing of the Wronskian of

two solutions does not depend on the choice of t.

} Therefore, we have the following theorem:

} Theorem 4: The Wronskian of the two solutions (2) of the

homogeneous system (1) is either identically zero or never
zero on I.

•18 •Anil Kumar, Dept. of Mathematics

Linear dependence/independence

The solutions í
ì x = x1 (t ) and ì x = x 2 (t )
}
í
î y = y1 (t ) î y = y 2 (t )
of the homogeneous system (1) are called linearly dependent on
I if x1(t) = k x2(t) or x2(t) = k x1(t)
y1(t) = k y2(t) y2(t) = k y1(t)
for some constant k and all t in I. Otherwise linearly
independent.
} Linear dependence is equivalent to the condition that there exist
two constants, A and B, and at least one of them is not zero, such
that Ax1(t) + B x2(t) = 0
Ay1(t) + B y2(t) = 0 for all t in I.
•Anil Kumar, Dept. of Mathematics •19

Linear dependence/independence

Theorem 5: If the two solutions (2) of the homogeneous

system (1) are linearly independent on I, then (3) is the
general solution of (1) on I.

Proof: To prove the theorem, it suffices to show that the solutions

(2) are linearly dependent iff their Wronskian W(t) is zero.

•20 •Anil Kumar, Dept. of Mathematics

Homogeneous linear systems with constant
coefficients
} Find a general solution of the following homogeneous system of
differential equations
ì dx
ïï dt = a1 x + b1 y
í (1)
ï dy = a x + b y
ïî dt 2 2

where a1, b1, a2, b2 are given constants.

} One method of finding the general solution of above system is
based on differentiating one of the two equations given,
eliminating one of the dependent variables and solving the
resulting second order linear equation.

•Anil Kumar, Dept. of Mathematics •21

EXERCISES
} By differentiating an equation of the system w.r.t. t and
eliminating any one of the dependent variable find the
general solution of the following
ì dx ì dx
ïï dt = x + 2 y ïï dt = x + y
(1) í (2) í
ï dy = 3x + 2 y ï dy = y
ïî dt ïî dt

Any 2nd order equation obtained from the system is

ì dx not equivalent to the system in the sense that it has
ïï dt = x solutions that are not part of any solution of the
(3) system.
í
ï dy = y
ïî dt Although, higher order equations are equivalent to the
systems, the reverse is not true and systems are more
general.
•Anil Kumar, Dept. of Mathematics •22
 Homogeneous linear systems
Method 2:
} We know that the exponential function has the property that its
derivatives are constant multiples of the function itself. Hence it is
natural to seek solution of (1) of the type x = Ae ü
mt

ý (2)
y = Bemt þ
Substituting in (1) we get
Amemt = a1 Aemt + b1 Bemt and Bmemt = a2 Aemt + b2 Be mt
and therefore, we get
(a1 - m) A + b1B = 0 ü
ý (3)
a2 A + (b2 - m) B = 0þ

•Anil Kumar, Dept. of Mathematics •23

Homogeneous linear systems

} The above equations have nontrivial solutions whenever

a1 - m b1
=0 Þ m2 - (a1 + b2 )m + (a1b2 - a2b1 ) = 0
a2 b2 - m

} This equation is called auxiliary equation of the system (1).

} Let m1 and m2 be the roots of the auxiliary equation.
} When m = m1, system (3) gives solutions, say A = A1 and B = B1.
} And then the corresponding nontrivial solution of system (1) is
ì x = A1e m1t
í
î y = B1e
m1t

•Anil Kumar, Dept. of Mathematics •24

Homogeneous linear systems
} When m = m2, system (3) gives solutions, say, A = A2 and B = B2.
} Then the corresponding nontrivial solution of system (1) is
x = A2em2t ü
ý
y = B2em2t þ
} Thus, two solutions are x = A em1t ü x = A2em2t ü
1
m1t ý
and ý
y = B1e þ y = B2em2t þ
} The above two solutions, in general, may not be linearly
independent.
} Therefore, while constructing the general solution, three cases on
the nature of the roots m1 and m2 of the auxiliary equation must
be dealt with carefully. (Three cases)
•Anil Kumar, Dept. of Mathematics •25

Case 1: m1 ≠ m2 and real

} When m1 ≠ m2 clearly the above set of solutions are linearly
independent.
} Hence the general solution is given by

x = c1 A1em1t + c2 A2em2t ü In this case the two solutions are

m2t ý linearly independent?
y = c1B1e + c2 B2e þ
m1t

} Example: Solve the system of linear first order equations

dx ü dx ü
(i ) = -3x + 4 y ï (ii ) = 4x - 3yï
dt ï dt ï
ý ý
dy dy
= -2 x + 3 y ï = 8x - 6 y ï
dt ïþ dt ïþ
•Anil Kumar, Dept. of Mathematics •26
Case 2: m1 = m2
} In this case, we get a repeated solution as

ì x = Ae mt
í
î y = Be
mt

} Get the second linearly independent solution as

x = ( A1 + A2t )e mt and y = ( B1 + B2t )e mt
and hence the general solution is
x = c1 ( A1 + A2t )e mt + c2 Ae mt
y = c1 ( B1 + B2t )e mt + c2 Be mt (*)

where c1 and c2 are arbitrary constants.

•Anil Kumar, Dept. of Mathematics •27

Examples

dx dx
= -4 x - y = 5x + 4 y
dt dt
(i ) (ii )
dy dy
= x - 2y = -x + y
dt dt

•Anil Kumar, Dept. of Mathematics •28

 Remark
} In formula (*), the constants A2 and B2 satisfy the same
linear algebraic system as the constants A and B, and
consequently, we may put
A2 = A and B2 = B
without loss of generality.

Therefore, the general solution of the system is

x = c1 ( A1 + At )e mt + c2 Ae mt
y = c1 ( B1 + Bt )e mt + c2 Be mt

•29 •Anil Kumar, Dept. of Mathematics

Case 3: m1, m2 are complex

} Let m = a ± ib, (where a, b are real numbers and b ≠ 0)
} For m1 = a + ib, the expected solution is of the type

ì x = ( A1 + iA2 )e( a +ib )t x = ( A1 + iA2 )e at (cos bt + i sin bt )

í ( a +ib ) t
Þ
î y = ( B1 + iB2 )e y = ( B1 + iB2 )eat (cos bt + i sin bt )

where A1, A2, B1, B2 are real constants such that neither both
of A1, A2, nor both of B1, B2 are zero. Therefore, we have
ì x = e at ( A1 cos bt - A2 sin bt ) + i( A1 sin bt + A2 cos bt )
í
î y = e ( B1 cos bt - B2 sin bt ) + i( B1 sin bt + B2 cos bt )
at

•30 •Anil Kumar, Dept. of Mathematics

Case 3: m1, m2 are complex…
} Easy to see that if a pair of complex-valued functions is a
solution of the system, in which the coefficients are real
constants, then their two real parts and their two
imaginary parts are real-valued solutions.
} From this fact, two real-valued solutions are
ì x = e at ( A1 cos bt - A2 sin bt ) ì x = eat ( A1 sin bt + A2 cos bt )
í and í
= - î y = e ( B1 sin bt + B2 cos bt )
at at
î y e ( B1 cos bt B 2 sin bt )
} Therefore, the general solution of the system is
x = c1eat ( A1 cos bt - A2 sin bt ) + c2eat ( A1 sin bt + A2 cos bt )
y = c1eat ( B1 cos bt - B2 sin bt ) + c2eat ( B1 sin bt + B2 cos bt )

•31 •Anil Kumar, Dept. of Mathematics

Case 3: m1, m2 are complex…

} Two solutions give the following pair of real solutions.

ì x = e at ( A1 cos bt - A2 sin bt ) ì x = eat ( A1 sin bt + A2 cos bt )

í and í
= - î y = e ( B1 sin bt + B2 cos bt )
at at
î y e ( B1 cos bt B 2 sin bt )
} Wronskian of the two solutions is given by

W (t ) = ( A1 B2 - A2 B1 )e 2 at ¹ 0 as A1 B2 - A2 B1 ¹ 0

and hence these two solutions are linearly independent.

•Anil Kumar, Dept. of Mathematics •32

Example

dx dx
= x - 2y = 4x - 2 y
dt dt
(i ) (ii )
dy dy
= 4x + 2 y = 5x + 2 y
dt dt

•Anil Kumar, Dept. of Mathematics •33

General solution of non homogeneous system of

equations.
ì x = Ax1 (t ) + Bx 2 (t )
Theorem 6: If í is the general solution of the
î y = Ay1 (t ) + By 2 (t )
homogeneous part of the non- homogeneous system
dx
= a1 (t ) x + b1 (t ) y + f1 (t )
dt
(1)
dy
= a2 (t ) x + b2 (t ) y + f 2 (t )
dt

on I and if x = x p (t ) and y = y p (t ) is any particular solution of

(1) on I, then ìí x = Ax1 (t ) + Bx 2 (t ) + x p (t ) is the general solution of the
î y = Ay1 (t ) + By 2 (t ) + y p (t )
system (1).
•34 •Anil Kumar, Dept. of Mathematics
 Exercise: Solution to non-homogeneous
system
ì dx
ïï dt = a1 (t ) x + b1 (t ) y + f1 (t )
} Consider a non-homogeneous linear system í
ï dy = a (t ) x + b (t ) y + f (t )
ïî dt 2 2 2

ì dx
ïï dt = a1 (t ) x + b1 (t ) y
and the corresponding homogeneous system í
ï dy = a (t ) x + b (t ) y
ïî dt 2 2

ì x = x1 (t ) ì x = x2 (t )
If í y = y (t ) and í are linearly independent solution
î 1 î y = y2 (t )
of homogeneous system so that

•Anil Kumar, Dept. of Mathematics •35

Exercise: Solution to non-homogeneous

system
ì x = Ax1 (t ) + Bx 2 (t )
í
î y = Ay1 (t ) + By 2 (t ) is its general solution, show that
ì x = v1 (t ) x1 (t ) + v2 (t ) x2 (t )
í
î y = v1 (t ) y1 (t ) + v2 (t ) y2 (t )
will be a particular solution of non-homogeneous system if the
functions v1(t) and v2(t) satisfy the system

v1/ x1 + v2/ x2 = f1
v1/ y1 + v 2/ y 2 = f 2
(Proof is Just the verification)
•Anil Kumar, Dept. of Mathematics •36
 Just the verification
} Note: This method of getting the particular solution of a non-
homogeneous system is called method of variation of
parameters

} Example: Use method of variation of parameters to find the

particular solution of the non homogeneous system
dx
= x + y - 5t + 2
dt
dy
= 4 x - 2 y - 8t - 8
dt
Also find the general solution.

•Anil Kumar, Dept. of Mathematics •37

Example
} Use the method of undetermined coefficients to find the
particular solution of the non-homogeneous system
dx
= x + y - 5t + 2
dt
dy
= 4 x - 2 y - 8t - 8
dt
Also, find the general solution.
Ø Consider the particular solution of the above system as
!! (#) = &1 # + (1
*! # = &2 # + (2

•38 •Anil Kumar, Dept. of Mathematics