INTRODUCTION TO BERNOULLI TECHNIQUE OF FIRST ORDER

DIFFERENTIAL EQUATION
WRITTEN AND EDITED BY
ADEBAYO YUSUFF T.
[MASTER KHAN 500LEVEL]
DEPARTMENT OF MATHEMATICAL SCIENCES
SCHOOL OF PHYSICAL SCIENCES
FEDERAL UNIVERSITY OF TECHNOLOGY AKURE [FUTA].

Bernoulli equation is non-linear differential equation of the

form
𝑑𝑦
+ 𝑝(𝑡)𝑦 = 𝑞(𝑡)𝑦 𝑛 … 𝑒𝑞𝑛1
𝑑𝑡
Where 𝑛 ∈ ℛ, 𝒏 ≠ 𝟎 𝒂𝒏𝒅 𝒏 ≠ 𝟏. The above equation is non-
linear in 𝑦(𝑡) but it can be reduced to an equation that is
linear in the new variable 𝑧(𝑡) using the transformation
technique. Suppose we divide above equation through by
𝑦 𝑛 and we have
𝑑𝑦
𝑦 −𝑛 + 𝑝(𝑡)𝑦1−𝑛 = 𝑞 (𝑡) … 𝑒𝑞𝑛2
𝑑𝑡

1−𝑛
𝑑𝑧 −𝑛
𝑑𝑦 𝑑𝑦 𝑦 𝑛 𝑑𝑧
𝑙𝑒𝑡 𝑧 = 𝑦 , = (1 − 𝑛)𝑦 𝑠𝑜 𝑡ℎ𝑒𝑟𝑒𝑓𝑜𝑟𝑒 =
𝑑𝑡 𝑑𝑡 𝑑𝑡 1 − 𝑛 𝑑𝑡
𝑑𝑦 𝑦 𝑛 𝑑𝑧
Replace 𝑏𝑦 𝑎𝑛𝑑 𝑦1−𝑛 𝑏𝑦 𝑧 𝑖𝑛 𝑒𝑞𝑢𝑎𝑡𝑖𝑜𝑛 2. 𝑤𝑒 𝑡ℎ𝑒𝑛 ℎ𝑎𝑣𝑒
𝑑𝑡 1−𝑛 𝑑𝑡
 −𝑛
𝑦 𝑛 𝑑𝑧
𝑦 + 𝑝(𝑡)𝑧 = 𝑞 (𝑡) … 𝑒𝑞𝑛3
1 − 𝑛 𝑑𝑡
So what we have next is equation that similar to linear
equation
1 𝑑𝑧
+ 𝑝(𝑡)𝑧 = 𝑞(𝑡)
1 − 𝑛 𝑑𝑡
Multiply through by 1 − 𝑛 and make it linear completely, we
have
𝑑𝑧
+ (1 − 𝑛)𝑝(𝑡)𝑧 = (1 − 𝑛)𝑞 (𝑡) … 𝑒𝑞𝑛4
𝑑𝑡
By letting 𝑃(𝑡) = (1 − 𝑛)𝑝(𝑡) 𝑎𝑛𝑑 𝑄 (𝑡) = (1 − 𝑛)𝑞(𝑡)

The above equation4 then becomes linear equation of the

form

𝑑𝑧
+ 𝑃(𝑡)𝑧 = 𝑄(𝑡) … 𝑒𝑞𝑛5
𝑑𝑡
And it can be solved by integrating factor method we learnt
in our last class.

𝑧. 𝐼𝐹 = ∫ 𝑄(𝑡) . 𝐼𝐹

Where

𝐼𝐹 = 𝑒 ∫ 𝑃(𝑡) .
Likely question: given the Bernoulli of degree 𝑛;
𝑑𝑦
+ 𝑝(𝑡)𝑦 = 𝑞 (𝑡)𝑦 𝑛 .
𝑑𝑡
Proof that the linearized version of the Bernoulli equation is
𝑑𝑢
+ (1 − 𝑛)𝑝(𝑡)𝑢 = (1 − 𝑛)𝑞 (𝑡) ℎ𝑖𝑛𝑡: 𝑢 = 𝑦1−𝑛
𝑑𝑡
𝑑𝑦
Using above, solve the equation 3(1 + 𝑡 2 ) = 2𝑡𝑦(𝑦 3 − 1)
𝑑𝑡

Solution: all you need to do in the first place is to convert

the above problem to the standard form of Bernoulli which
is:
𝑑𝑦
+ 𝑝(𝑡)𝑦 = 𝑞 (𝑡)𝑦 𝑛
𝑑𝑡
𝑑𝑦
3(1 + 𝑡 2 ) = 2𝑡𝑦(𝑦 3 − 1)
𝑑𝑡
Can be written by open the bracket on the 𝑟ℎ𝑠 as
𝑑𝑦
3(1 + 𝑡 2 ) = 2𝑡𝑦 4 − 2𝑡𝑦
𝑑𝑡
𝑑𝑦 2𝑡 2𝑡
+ 2
𝑦 = 2
𝑦 4 … 𝑒𝑞𝑛1
𝑑𝑡 3(1 + 𝑡 ) 3(1 + 𝑡 )
2𝑡 2𝑡
Now 𝑝(𝑡) = , 𝑞 (𝑡) = 𝑎𝑛𝑑 𝑦 𝑛 = 𝑦 4
3(1+𝑡 2 ) 3(1+𝑡 2 )

Divide through by 𝑦 4 we have

 1 𝑑𝑦 2𝑡 𝟏 2𝑡
+ = … 𝑒𝑞𝑛 2
𝑦 4 𝑑𝑡 3(1 + 𝑡 2 ) 𝒚𝟑 3(1 + 𝑡 2 )
How do u know the variable to use in transforming?
Always choose variable 𝑦 which is coefficient of 𝑝(𝑡)
𝟏 𝑑𝑧 𝟑 𝒅𝒚
𝑧= , =− 𝟒
𝒚𝟑 𝑑𝑡 𝒚 𝒅𝒕
𝑑𝑦
Making subject of formula then we have the following
𝑑𝑡

𝑑𝑦 −𝒚𝟒 𝒅𝒛
=
𝑑𝑡 𝟑 𝒅𝒕
𝑑𝑦 −𝒚𝟒 𝒅𝒛 𝟏
Go back to eqn2 and sub = 𝒂𝒏𝒅 𝑧 =
𝑑𝑡 𝟑 𝒅𝒕 𝒚𝟑

1 −𝒚𝟒 𝒅𝒛 2𝑡 2𝑡
( ) + 𝒛 =
𝑦 4 𝟑 𝒅𝒕 3(1 + 𝑡 2 ) 3(1 + 𝑡 2 )

Doing the normal mathematical operations here we have

𝑑𝑧 2𝑡 2𝑡
− 𝑧 = −
𝑑𝑡 1 + 𝑡 2 1 + 𝑡2
Now non-linear equation has been transformed to linear
𝑑𝑦
equation of the form + 𝑝(𝑡)𝑦 = 𝑞 (𝑡) but now in term of 𝑧
𝑑𝑡

And it can be solved using integrating factor techniques

2𝑡 2𝑡
Now 𝑝(𝑡) = − 𝑎𝑛𝑑 𝑞 (𝑡) = −
1+𝑡 2 1+𝑡 2

𝑖𝑛𝑡𝑒𝑔𝑟𝑎𝑡𝑖𝑛𝑔 𝑓𝑎𝑐𝑡𝑜𝑟 𝜇 (𝑡) = 𝑒 ∫ 𝑝(𝑡)

2𝑡
∫ 𝑝(𝑡) = − ∫
1 + 𝑡2
 −ln(1 + 𝑡 2 )
2) 2 )−1 1
𝜇 (𝑡) = 𝑒 −ln(1+𝑡 = 𝑒 𝑙𝑛(1+𝑡 =
1 + 𝑡2
1
𝜇(𝑡) =
1 + 𝑡2
Now the final answer is

𝑧𝜇 (𝑡) = ∫ 𝑞(𝑡) 𝜇 (𝑡)𝑑𝑡

1 2𝑡 1 2𝑡
𝑧 = ∫ − 𝑑𝑡 = ∫ − 𝑑𝑡
1 + 𝑡2 1 + 𝑡2 1 + 𝑡2 (1 + 𝑡 2 )2
integrating right hand side by method of substitution.
𝑑𝑢
𝑙𝑒𝑡 𝑢 = 1 + 𝑡 2 , 𝑑𝑢 = 2𝑡𝑑𝑡, 𝑑𝑡 =
2𝑡
1 2𝑡 𝑑𝑢 𝑑𝑢 1
𝑧 = ∫ = ∫ = − +𝑐
1 + 𝑡2 𝑢2 2𝑡 𝑢2 𝑢
1 1
𝑧 = − +𝑐
1 + 𝑡2 1 + 𝑡2
𝑧 = −1 + 𝑐(1 + 𝑡 2 )
1
Don’t forget that 𝑧 =
𝑦3

So
1
3
= −1 + 𝑐(1 + 𝑡 2 ) .
𝑦
 1
Example2: by means of substituting 𝑧 = , transform the
𝑦2
𝑑𝑦
differential equation + 𝑦 = 𝑥𝑦 3 into equation containing 𝑧
𝑑𝑥
and 𝑥. Solve this equation for 𝑧 and hence show that the
2
solution of the given problem is 𝑦 2 = 𝑔𝑖𝑣𝑒𝑛 𝑦 =
2𝑥+1
√2 𝑤ℎ𝑒𝑛 𝑥 = 0.
Solution:
𝑑𝑦
+ 𝑦 = 𝑥𝑦 3 … 𝑒𝑞𝑛1
𝑑𝑥
Divide through by 𝑦 3 we have the following
1 𝑑𝑦 𝑦
+ =𝑥
𝑦 3 𝑑𝑥 𝑦 3
1 𝑑𝑦 1
+ = 𝑥 … 𝑒𝑞𝑛2
𝑦 3 𝑑𝑥 𝑦 2
1 𝑑𝑧 2 𝑑𝑦
𝑙𝑒𝑡 𝑧 = , =− 3
𝑦2 𝑑𝑥 𝑦 𝑑𝑥
𝑑𝑦
Make subject of formula we have
𝑑𝑥

𝑑𝑦 𝑦 3 𝑑𝑧
=−
𝑑𝑥 2 𝑑𝑥
Sub
1 𝑑𝑦 𝑦 3 𝑑𝑧
𝑧= 2 =− 𝑖𝑛𝑡𝑜 𝑒𝑞𝑛 2
𝑦 𝑑𝑥 2 𝑑𝑥
1 𝑦 3 𝑑𝑧
(− )+𝑧=𝑥
𝑦3 2 𝑑𝑥
 𝑑𝑧
− 2𝑧 = −2𝑥
𝑑𝑥
Now non-linear equation has been transformed to linear
𝑑𝑦
equation of the form + 𝑝(𝑥 )𝑦 = 𝑞(𝑥 ) but now in term of 𝑧
𝑑𝑥

And it can be solved using integrating factor techniques

Now 𝑝(𝑥 ) = −2 𝑎𝑛𝑑 𝑞 (𝑥 ) = −2𝑥

𝑖𝑛𝑡𝑒𝑔𝑟𝑎𝑡𝑖𝑛𝑔 𝑓𝑎𝑐𝑡𝑜𝑟 𝜇 (𝑥 ) = 𝑒 ∫ 𝑝(𝑥)

∫ 𝑝(𝑥) = − ∫ 2

= −2𝑥
𝜇 (𝑥 ) = 𝑒 −2𝑥
Now the final answer is

𝑧𝜇 (𝑥 ) = ∫ 𝑞(𝑥) 𝜇 (𝑥 )𝑑𝑥

𝑧𝑒 −2𝑥 = ∫ −2𝑥𝑒 −2𝑥 𝑑𝑥

Integrating right hand side by part method

−2𝑥
𝑒 −2𝑥
𝑙𝑒𝑡 𝑢 = −2𝑥, 𝑑𝑢 = −2𝑑𝑥, 𝑑𝑣 = 𝑒 , 𝑣=−
2

𝑢𝑣 − ∫ 𝑣𝑑𝑢

𝑧𝑒 −2𝑥 = 𝑥𝑒 −2𝑥 − ∫ 𝑒 −2𝑥 𝑑𝑥

 −2𝑥 −2𝑥
𝑒 −2𝑥
𝑧𝑒 = 𝑥𝑒 + +𝑐
2
1
Don’t forget that 𝑧 =
𝑦2

So
1 −2𝑥 −2𝑥
𝑒 −2𝑥
𝑒 = 𝑥𝑒 + +𝑐
𝑦2 2
Divide through by 𝑒 −2𝑥
1 1
2
= 𝑥 + + 𝑐𝑒 2𝑥 .
𝑦 2
Now let impose initial condition given to us

𝑦 = √2 𝑤ℎ𝑒𝑛 𝑥 = 0
1 1
=0+ + 𝑐𝑒 2(0)
(√2)2 2
1 1
= +𝑐
2 2
So 𝑐 = 0 and when 𝑐 = 0 we have our final answer as
1 1
= 𝑥 +
𝑦2 2
Or
1 2𝑥 + 1
=
𝑦2 2
Making 𝑦 2 subject of formula
 2
𝑦2 =
2𝑥 + 1
That’s all!!!

One more point to take care of is not all the time will
equation be non-linear by power of 𝑛 > 1 like the two
previous examples, the equation can be non-linear by
function of 𝑦 sometimes.
𝑑𝑦
+ tan 𝑡 tan 𝑦 = 𝑐𝑜𝑠𝑡𝑠𝑒𝑐𝑦
𝑑𝑡
Is a non-linear problem with functions of 𝑦
Solution: the above problem can be written as the following
𝑑𝑦 sin 𝑦 1
+ tan 𝑡 = cos 𝑡 … 𝑒𝑞𝑛1
𝑑𝑡 cos 𝑦 cos 𝑦
Multiply through by 𝑐𝑜𝑠𝑦 we shall have the following
𝑑𝑦
cos 𝑦 + tan 𝑡 sin 𝑦 = cos 𝑡 … 𝑒𝑞𝑛2
𝑑𝑡
𝑙𝑒𝑡 𝑧 = sin 𝑦 Which is coefficient of 𝑝(𝑡) = tan 𝑡
𝑑𝑧 𝑑𝑦
= 𝑐𝑜𝑠𝑦
𝑑𝑡 𝑑𝑡
𝑑𝑦
Make subject of formula, we have
𝑑𝑡
𝑑𝑦 1 𝑑𝑧 𝑑𝑦
= after substituting 𝑧 𝑎𝑛𝑑 we have the following
𝑑𝑡 𝑐𝑜𝑠𝑦 𝑑𝑡 𝑑𝑡

1 𝑑𝑧
cos 𝑦 ( ) + z tan 𝑡 = cos 𝑡
𝑐𝑜𝑠𝑦 𝑑𝑡
Now non-linear equation has been transformed to linear
𝑑𝑦
equation of the form + 𝑝(𝑡)𝑦 = 𝑞 (𝑡) but now in term of 𝑧
𝑑𝑡

𝑑𝑧
+ 𝑧 𝑡𝑎𝑛𝑡 = 𝑐𝑜𝑠𝑡
𝑑𝑡
And it can be solved using integrating factor techniques
Now 𝑝(𝑡) = tan 𝑡 𝑎𝑛𝑑 𝑞 (𝑡) = cos 𝑡

𝑖𝑛𝑡𝑒𝑔𝑟𝑎𝑡𝑖𝑛𝑔 𝑓𝑎𝑐𝑡𝑜𝑟 𝜇 (𝑡) = 𝑒 ∫ 𝑝(𝑡)

sin 𝑡
∫ 𝑝(𝑡) = ∫ = −𝑙𝑛𝑐𝑜𝑠𝑡
cos 𝑡
−1 1
𝜇 (𝑡) = 𝑒 −ln cos 𝑡 = 𝑒 𝑙𝑛(cos 𝑡) =
cos 𝑡
1
𝜇 (𝑡) =
cos 𝑡
Now the final answer is

𝑧𝜇 (𝑡) = ∫ 𝑞(𝑡) 𝜇 (𝑡)𝑑𝑡

1 cos 𝑡
𝑧 =∫ 𝑑𝑡 = ∫ 𝑑𝑡
cos 𝑡 𝑐𝑜𝑠𝑡
Integrating right hand side we have
1
𝑧 =𝑡+𝑐
cos 𝑡
𝑧 = 𝑡 cos 𝑡 + 𝑐 cos 𝑡
Since 𝑧 = sin 𝑦 our final answer is therefore
𝑠𝑖𝑛𝑦 = 𝑡 cos 𝑡 + 𝑐 cos 𝑡
Remember I told you about general solution and particular
solution.
General solution is the solution that contain 𝑐 unknown value while
particular solution is the solution in which the value of 𝑐 is known
by imposing initial value condition.