Sujin Khomrutai, Ph.D.
Department of Math & Computer Science Chulalongkorn University
Lecture 1
1 Introduction
2 Some Aspects
3 First Order ODEs: A Review
4 Second Order Linear Equations
Definition
A differential equation(or DE) is an equation involving a function and some of its derivatives.
EX.
1 y0 = 2xy is a DE. (variable x, a function y(x).)
2 x2+u2 = 4 is nota DE. No derivative!
3 t2+ (y00)2 = 4 is a DE. (variablet, a function y(t).)
4 ux+ 5uy = 1 is a DE. (variablesx,y, a functionu(x,y).)
Definition
A system of DEsis a group of equations involving two or more functions and some of their derivatives.
EX.
1
(u0 = 2u+v
v0=−u+ 4v. ⇒ a system of DEs. (functions u,v)
2
(x+ 2y= 2
−x+y = 0. ⇒not a system of DEs. No derivatives!
3
u= 3u−v v0 =u+w w0=−1
⇒ a system of DES. (functions u,v,w)
Definition In a DE,
1 only one variable⇒ ordinary differential equation (or ODE).
2 two or more variables⇒ partial differential equations(or PDE).
System of ODEs = 1 variable, at least 2 functions
System of PDEs= at least 2 variables, at least 2 functions.
EX.
1 y0 = 2xy is an ODE.
2 uxx+uyy = 1 is a PDE.
3
(u0 = 2u+v
v0=−u+ 4v is a system of ODEs.
Goal. We study DEs and sytems of DEs.
Variable. t= time, or x= a space variable.
Functions. y (ODE), vectors~x= (x1, . . . ,xn) (Systems of ODES) Derivatives.
y0 = dy
dt, y00 = d2y
dt2, . . . ,y(n)= dny dtn, . . . and
~
x0 = (x10,x20, . . . ,xn0), ~x00= (x100,x200, . . . ,xn00), . . .
Definition (Order)
For a DE or system of DEs, the highest order of derivative appeared is called the order.
EX.
1 y00+ 2y0=t3 ⇒ order = 2. (ODE)
2 y(4)−y00+ (y0)5= 0 ⇒ order = 4. (ODE)
3
(x10 = 2x1−x2
x20 =−x1+x2
⇒ order = 1. (system of ODEs)
4 ut+u(uxxt)5 = 0⇒ order = 3. (PDE)
Definition (Solutions)
A solution to a DE is a function that the equation becomes true when we substitute the function and its derivatives into the equation.
EX.
1 y0 = 2t ⇒ y(t) =t2.
2 y0 = 2y ⇒ y(t) =e2t.
3 y0+ 3y = 6 ⇒y(t) =e−3t+ 2.
4 y00−3y0+ 2y = 0 ⇒ y(t) =C1et+C2e2t.
5 y(4) = 1 ⇒ y(t) = 24t4.
Definition (Particular & General solutions)
A solution to a DE that contains no arbitrary constants is called a particular solution.
A solution to a DE that contains some arbitrary constants and represents all possible particular solution is called ageneral solution.
EX.Consider the ODE y0 = 2t.
y1(t) =t2 and y2(t) =t2+ 3 are solutions. They are particular solutions.
To find a general solution, we need integration y0= 2t ⇒
Z
y0dt = Z
2tdt which gives
y(t)−y(0) =t2 ⇒ y(t) =t2+C. where C =y(0) is an arbitrary constant.
Definition (Linear vs Nonlinear)
An ODE is called alinear ODEif it can be expressed as any(n)+· · ·+a2y00+a1y0+a0y =f(t).
Otherwise, it is called nonlinear ODE.
EX.
1 y00+ 2ty0−3y =et ⇒linear.
2 ty(4)= 3y00+ sint ⇒ linear.
3 y0+y2 = 0⇒ nonlinear.
4 y000+y00+ 2y0+ 4y =x2 ⇒ linear. (Observe: variable isx.)
Definition (Initial Value Problems)
An initial value problem (or IVP) is a problem that has a DE together with some conditions for the solutions at some t0 (or somex0).
Solving IVP⇒ Particular solution.
EX.Solve the initial value problem
y0 = 3t+ 2, y(0) = 0.
Sol. We integrate
y0 = 3t+ 2 ⇒ Z
y0dt = Z
(3t+ 2)dt then
y(t)−y(0) = 3t2 2 + 2t Using y(0) = 0, we obtain
y(t) = 3t2 2 + 2t.
Definition
A first order ODE is an equation of the form F(t,y,y0) = 0.
If it can be put into the form
y0=f(t)g(y) it is calledseparable.
If it can be put into the form
y0+p(t)y =q(t) it is a linear equation.
How to solve separable eqns.?
To solve a separable equation, we write dy
dt =f(t)g(y) ⇒ 1
g(y)dy =f(t)dt.
Then integrate
Z 1 g(t)dy =
Z
f(t)dt.
Solution can be implicitly defined.
EX.Solve the ODE
dy dt = t
y. Sol. We write
ydy=tdt.
Integrate Z
ydy = Z
tdt ⇒ y(t)2
2 −y(0)2 2 = t2
2. Setting C =y(0)2, we get
y(t)2=t2+C. This is a general solution.
EX.Sole the ODE
dy
dx = 3x2 y−siny. Sol. We express
dy
dx = 3x2
y−siny ⇒ (y−siny)dy = 3x2dx.
Then integrate Z
(y−siny)dy = Z
3x2dx ⇒ y2
2 + cosy =x3+C. The solution is implicit!
EX.Sole the ODE
y0 =xy. Sol. We write
y0 =xy ⇒ 1
ydy =xdx.
Integrate
Z 1 ydy =
Z
xdx ⇒ lny(t) = x2 2 +C. So
y(t) =ex
2
2+C =Dex
2 2.
Definition
For a linear equation
y0+p(t)y =q(t), theintegrating factor is the function
I(t) =e
Rp(t)dt.
Multiplying the ODE with I(t) and integrating, then y(t) = 1
I(t) Z
I(t)q(t)dt+C
.
EX.Solve the ODE
y0+ 3t2y = 6t2.
Sol. This is a linear equation: p(t) = 3t2 andq(t) = 6t2. The integrating factor
I(t) = exp Z
3t2dt =et3. Plug into the formula
y(t) = 1 et3
Z
et36t2dt+C
∴ y(t) =e−t3
2et3+C
= 2 +Ce−t3.
x2y0+xy = 1 (x>0), y(1) = 2.
Sol. Variable isx! Divide by x2 to bring coefficient of y0 to 1:
y0+ 1 xy = 1
x2.
This is a linear equation: p(x) = 1/x and q(x) = 1/x2. Integrating factor
I(x) = exp Z 1
xdx =elnx =x.
Plug into the formula y(x) = 1
x Z
x· 1
x2dx+C
= lnx+C
x .
Definition
A second order ODE is an equation of the form F(t,y,y0,y00) = 0.
If the ODE can be put into the form
a(t)y00+b(t)y0+c(t)y =f(t) or
y00+p(t)y0+q(t)y =r(t) it is called a second order linear ODE.
Ifa,b,c are constants, the ODE is calledconstant coefficients.
Iff(t) = 0 or r(r) = 0, the ODE is calledhomogeneous.
EX.Consider the ODE
3y00+ 2y0+ 4y=t2+ 1.
This is a second order ODE.
It is linear, constant coefficients, and nonhomogeneous.
EX.Consider the ODE
5y00+ (tant)y0+y2= 0.
This is a second order ODE.
It is nonlinear because of the term y2.
EX.Consider the ODE
ty00+ (t+ 1)y0+t2y = 0.
This is a second order ODE
a(t) =t, b(t) =t+ 1, c(t) =t2, f(t) = 0.
It is linear, non-constant coefficients, and homogeneous.
We can divide by t to get another form y00+ t+ 1
t y0+ty = 0.
So p(t) = (t+ 1)/t andq(t) =t.
Definition
If a second order ODE is supplied with two conditions of the form y(t0) =k, y0(t0) =l
it is called an initial value problem (or IVP).
Thus an IVP for a second order linear ODE is (IVP)
(y00+p(t)y0+q(t)y =f(t) y(t0) =k, y0(t0) =l. Usually but not always, we choose t0 = 0.
Theorem
Consider an IVP of a second order linear ODE (IVP)
(y00+p(t)y0+q(t)y =f(t) y(t0) =k, y0(t0) =l.
If p(t),q(t),r(t)are continuous functions on an interval I , then the IVP has a solution and it is unique.
The solution depends on the initial conditions k,l .
• The theorem can be applied to determine the longest intervalI so that the problem has a unique solution.
(IVP)
((t−1)y00+ty0+ (t+ 1)y = 0 y(2) = 2, y0(2) = 3.
has a unique solution.
Sol. The conclusion does not involve the initial conditions!
Divide by (t−1) so the coefficient of y00 becomes 1 y00+ t
t−1y0+t+ 1 t−1y = 0.
The functions p(t) =t/(t−1),q(t) = (t+ 1)/(t−1), and r(t) = 0 are continuous on (−∞,1)∪(1,∞).
The interval containing the initial time t0 = 2 is (1,∞) so the longest
EX.Find the solution to the IVP (IVP)
(y00+p(t)y0+q(t)y= 0 y(t0) = 0, y0(t0) = 0.
where p(t),q(t) are continuous functions.
Sol. By the theorem there is a unique solution.
Note that if we substitutey(t) = 0, it works
000+p(t)00+q(t)0 = 0, 0(t0) = 0, 00(t0) = 0.
So we get the solution to the IVP to bey(t) = 0.
Theorem
If y1,y2 are solutions of a homogeneous ODE y00+p(t)y0+q(t)y = 0, then C1y1+C2y2 is also a solution.
Note. It is not true for nonhomogeneous equation y00+p(t)y0+q(t)y =r(t).
In order thatC1y1+C2y2 is a general solution y1,y2 have to be extra!
EX.Consider the ODE
y00+ 3y0+ 2y= 0.
It can be shown that
y1(t) =e−t, y2(t) =e−2t
are solutions. Using the superposition principle, we form a solution formula y(t) =C1e−t+C2e−2t.
Whether this is a general solution depends on whether it can satisfy the initial conditions
y(t0) =k, y0(t0) =l, for anyk,l.
Plug the formula y(t) into the conditions
(C1y1(t0) +C2y2(t0) =k, C1y10(t0) +C2y20(t0) =l.
This system can be solved for any k,l if and only if det
y1(t0) y2(t0) y10(t0) y20(t0)
6= 0.
Definition
For any two functions f,g, theWronskian of f,g is the function W(f,g) = det
f(t) g(t) f0(t) g0(t)
.