Scribd
Upload
70 views16 pages

Schap 22

This document discusses techniques for solving nonlinear first-order differential equations. It begins by introducing autonomous and non-autonomous equations, and describes how to draw and analyze phase diagrams to understand qualitative behavior. It then discusses using Bernoulli's method and separation of variables to solve certain types of non-linear first-order equations. The document also applies these concepts to the Solow growth model with and without technological progress.

Uploaded by

Aljebre Mohmed
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
70 views16 pages

Schap 22

This document discusses techniques for solving nonlinear first-order differential equations. It begins by introducing autonomous and non-autonomous equations, and describes how to draw and analyze phase diagrams to understand qualitative behavior. It then discusses using Bernoulli's method and separation of variables to solve certain types of non-linear first-order equations. The document also applies these concepts to the Solow growth model with and without technological progress.

Uploaded by

Aljebre Mohmed
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 16

Chapter 22

Non-Linear, First-Order Differential Equations

In this chapter, we will learn:


1. How to solve nonlinear first-order dif-
ferential equation?
2. Use of phase diagram in order to under-
stand qualitative behavior of differential
equation.

Autonomous Differential Equation

The initial-value problem for an autonomous,


nonlinear, first-order differential equation has
the following form:

ẏ = g(y(t)) & y(t0 ) = y0 (22.1)

dg(y) d2 g(y)
where dy & dy2 6= 0.

1
Phase Diagram

Although, it is known that solution to


(22.1) exists under the condition that dg(y) dy
is continuous in the neighborhood around t0 ,
in most cases it is not possible to derive the
explicit solution. Often qualitative proper-
ties of the differential equation are derived by
plotting it. Such plots are known as phase
diagram.

Steps in Drawing Phase Diagram


Let the differential equation

ẏ = g(y(t)). (22.2)
Our goal is to plot ẏ or g(y(t)).

Step 1 Take ẏ or g(y(t)) on y-axis and y(t) on


x-axis.

2
Step 2 Take the first and second derivative of ẏ
or g(y(t)) with respect to y, dg(y)
dy and
d2 g(y)
dy 2 .This gives you the shape of the
curve (increasing, decreasing, concave, c-
onvex).

Step 3 Derive the steady-state points by setting

ẏ = g(y(t)) = 0. (22.3)
Steady-state or equilibrium points are t-
he points at which the curve of ẏ or g(y)
intersects the x-axis. There can be more
than one steady-state point (multiplicity
of equilibria).

Given that there can be multiplicity of


equilibria, it raises the question which steady-
state points are stable and which are unsta-
ble.

3
Stability Analysis

Stability analysis tells us about the con-


vergence property of the differential equation.
A steady state point is stable, if the differen-
tial system converges to that point. Other-
wise, it is unstable.

Theorem 22.2: A steady-state equilib-


rium point of a nonlinear first-order dif-
ferential equation is stable if the deriva-
tive ddyẏ < 0 at that point and unstable if
the derivative is positive at that point

4
Solow or Neo-Classical Growth Model

The model relates long run per-worker con-


sumption and growth rate in output to sav-
ing rate, work-force growth rate, and techni-
cal progress.

Assumptions

1. Constant returns to scale production te-


chnology, Y = F (K, L)
2. Diminishing Marginal Productivity of C-
apital
3. Constant Rate of saving (s), thus total
savings is S(t) = sY (t),
4. Constant labor force growth rate (n),
5. Constant depreciation rate (δ).
6. No technical progress (Temporary Ass-
umption)

5
Implications of the Model

1. In the long run economy reaches a stable


steady state equilibrium.

2. Per-Worker Consumption (c) in the long


run depends on s, n, and δ. There will
be no growth in c in the long run.

3. Ultimately, an economy will grow at the


rate of work-force growth (n).

Let I(t) be gross investment, then by def-


inition, growth rate of capital stocks is

K̇ = I(t) − δK(t).
Since in equilibrium S(t) = I(t), we have

K̇ = sY (t) − δK(t).
Now define capital-labor ratio as, κ = K
L.
Then, given constant returns to scale, per-
worker output, y(t) can be written as

6
Y (t) K(t)
y(t) = = F( , 1) = f (κ(t)).
L(t) L(t)

Also

K̇ sY (t) − δK(t)
= = sy(t) − δκ(t).
L(t) L(t)

K̇ L̇
κ̇ = −κ
L(t) L(t)
Combining the above two equations, we get
differential equation in the capital-labor ratio
wh- ich characterizes Solow growth model.

κ̇ = sf (κ(t)) − (δ + n)κ(t).

7
Nonautonomous and Nonlinear Equation

The general form of the nonautonomous, fi-


rst-order differential equation is

ẏ = f (t, y). (22.5)

The equation can be a nonlinear function of


both y and t. We will consider two classes of
such equations for which solutions can be eas-
ily found: Bernoulli’s Equation and Sep-
arable Equations.

Bernoulli’s Equation

The differential equation

ẏ + a(t)y = b(t)y n , n 6= 0 or 1 (22.6)

is known as Bernoulli’s Equation. Assume


that a(t) and b(t) are continuous on some

8
time interval T . Now we can transform (22.6)
as follows:

Step 1 Multiply both sides of (22.6) by y −n . We


end up with

y −n ẏ + a(t)y 1−n = b(t). (22.7)

Step 2 Now define a new variable x = y 1−n .


Taking the derivative of x with respect
to time t, we get

ẋ = (1 − n)y −n ẏ. (22.8)

Step 3 Using the definition in step 2, differential


equation (22.7) can be written as

ẋ + (1 − n)a(t)x = (1 − n)b(t). (22.9)

9
This is first-order linear nonautonomo-
us differential equation, which can be so-
lved by using techniques learned in the
previous chapter.

Step 4 Once we have solved for x(t), we make


use of definition in step 2, x = y 1−n and
derive the solution for y(t).

Remark: This procedure is valid only when


y(t) 6= 0 forall t ∈ T .

10
Solow Growth Model With Technical Change

Earlier, we considered Solow growth mo-


del without technical change. Now, we in-
troduce technical change. Now suppose that
production depends on capital, k, and effec-
tive labor, EL defined as

EL(t) = E(t)L(t)
where E(t) is a measure of technology. Sup-
pose E(t) evolves as follows

Ė = λE(t), λ > 0.
Such technical change is called labor aug-
menting. Production function is

Y (t) = F (K(t), EL(t)) = K(t)α EL(t)1−α .

Rest of the model is identical to the previous


one. Now define capital-effective labor ratio

11
K
as, κ = EL . Then, given constant returns to
scale, output per effective labor unit, y(t) can
be written as

Y (t) K(t)
y(t) = = F( , 1) = κ(t)α .
EL(t) EL(t)

Also

K̇ sY (t) − δK(t)
= = sy(t) − δκ(t).
EL(t) EL(t)

K̇ ĖL
κ̇ = −κ
EL(t) EL(t)
Combining the above two equations, we get
differential equation in capital-effective labor
ratio which characterizes Solow growth model
with technical change.

κ̇ + (δ + λ + n)κ(t) = sκ(t)α .

12
As we can see that it is a Bernoulli equation.
In order to solve this multiply both sides by
κ−α and define x = κ1−α . Then the above
equation can be transformed into

ẋ + (δ + λ + n)(1 − α)x(t) = (1 − α)s.


The solution is

s
x(t) = + C exp−(δ+λ+n)(1−α)t .
δ+λ+n
In terms of κ(t), we get

κ(t) =
· ¸ 1−α
1
s
+ C exp−(δ+λ+n)(1−α)t .
δ+λ+n
Steady-state κ(t) is given by
· ¸ 1−α
1
s
κ= .
δ+λ+n

13
Implications

1. Ultimately economy reaches steady-state


just as in case of no technical progress.
2. At the steady state output, Y , consump-
tion, C, and capital stock, K, grow at
the rate of λ + n.
3. At the steady state per-worker output,
Y C
L , consumption per worker, L , and cap-
ital stock per worker, K
L , grow at the rate
of λ.

14
Separable Equations

The nonautonomous equation is

ẏ = f (t, y). (22.10)


f (t, y) can always be written as the ratio of
two other functions, M (t, y), and −N (t, y).
We can then rewrite (22.10) as

M (t, y) + N (t, y)ẏ = 0. (22.11


Definition : A non-linear, first-order
differential equation is separable if
M (t, y) = A(t), a function of only t, and
N (t, y) = y, a function of only y. A sep-
arable, nonlinear, first-order differential
equation can therefore be written as

A(t) + b(y)ẏ = 0. (22.12)

(22.12) can be solved by direct integration.


(22.12) can be written as

15
A(t)dt + b(y)dy = 0. (22.13)

This equation can be integrated directly to


obtain

Z Z
A(t)dt + b(y)dy = C. (22.14)

16

You might also like