Everything in Differential Equations

Great study guide for this class

Made by Jack Young (I want credit)
If you guys need help with any calc questions or need me to explain
something in these notes, then text me -> my number is (786) 574-
0024
Yes, I made flashcards for Calc 1 and Calc 2
You can find them all on this link: https://tinyurl.com/jackscalcnotes

If there is a Review
In the top right of a page, it just
means that is from Calc 1/2
Formulas that might be useful Review
Important Trig Double Angle Even/Odd Angle Properties 𝐥𝐧 𝒙 Rules

Addition & Subtraction

Special Right Triangles

Pythagorean
Unit Circle

Power Reduction Product-to-sum

Derivative Formulas Review
Implicit Differentiation
1. Differentiate both sides For second derivative,
take the derivative
2. When you take the implicitly again and
derivative replace with the first
of a y, multiply it by derivative
3. Solve for
Integration Formulas
ALSO CALLED ANTIDERIVATIVES
“When in doubt, do the opposite of derivatives” - Jack 2023
Review

This space used to be

for separation of
Chain Rule variables to solve a DE
- let u = inner function
- find du
- replace all x terms with u terms Now that’s the entire
- integrate basis of this class.
- put x back
More Integration Formulas Review
This space is just a sign to tell you to
take a quick break because oh my
god trig substitution is the worst.

Integration by parts Powers of sin & cos

n odd
Make u something that gets simpler as you take derivative of it
• L (log/ln)
• I (inverse trig) Make dv something that
• P (polynomial) you can integrate easily n even
• E (exponential)
• T (trig)

Indeterminate Forms m odd

When a bound has ∞ in it,
replace it with a variable and use n odd
a limit to solve the integral.
• If the final answer has ∞ in it, it
diverges m & n even
• If the final answer is a number,
it converges

Partial Fractions Powers of sec & tan

m odd

n even

m even or & n odd

Trig-Substitution
 Linear vs Nonlinear
The Basics A linear DE is linear if it follows these rules and
nonlinear if it doesn’t:
Previous Knowledge
• It needs to be able to be written in this form:
You should know by now all the trig rules, log/ln rules,
𝒅𝒏𝒚 𝒅𝒚
derivative formulas, integral formulas, and how to do 𝒂𝒏 𝒙 + 𝒂 𝟏 𝒙 + 𝒂𝟎 𝒙 𝒚 = 𝒈(𝒙)
partial derivatives. If you don’t remember, they are all 𝒅𝒙𝒏 𝒅𝒙
included before this slide. “n” can be anything, and there can be many
different n terms (like 2, 3, etc). The different a(x)’s
What’s a Differential Equation? are just different functions that involve x.
It is an equation that involves the derivative of a variable with • In that one term 𝑎0 𝑥 𝑦,There has to be that y there.
respect to another. 𝑑𝑦 • The y cannot be sin 𝑦 , 𝑒 𝑦 , or any other function of y it
• The first variable is the dependent variable 𝑑𝑥
− 3𝑦 = 0 needs to just be y.
• The second variable is the independent variable y is the dependent • None of the derivatives can have exponents
x is the independent • Each derivative needs to be its own term (no
Types of Differential Equations derivatives multiplied by each other)
An ordinary differential equation (ODE) only • The whole equation should only involve 1 dependent
involves derivatives with respect to one variable that is independent on 1 independent
independent variable. Only ODEs are variable
2
𝑑 𝑦 𝑑𝑦 𝜕𝑢 𝜕𝑢
𝑑𝑥 2 −2
𝑑𝑥
+ 6𝑦 = 0 𝜕𝑦
=−
𝜕𝑥
in this class! Explicit vs Implicit Solutions
An explicit solution makes the dependent variable
A partial differential equation (PDE) involves partial
alone and equal to an equation with independent
derivatives of more than one independent variable
variables. 1
ex. 𝑦 = 3 𝑥 2 + 8𝑥 − 𝑒 𝑥
What is the order of a DE? 3
(Usually in the form 𝑓(𝑥) = 𝑦)
𝑑2 𝑦 𝑑𝑦
The order is just the highest-order derivative in the 𝑑𝑥 2 + 3 𝑑𝑥 = 𝑒𝑥 An implicit solution mixes the independent and
equation. It is NOT the highest-order exponent! The order of this DE is 2 dependent variables on one (or both) side(s) of the
solution.
Let’s do an example! 4 5
(Usually in the form 𝑓(𝑥, 𝑦) = 0) ex. 5 𝑦2 − 𝑥3 − 8𝑥𝑦 = 0
Classify the 3 3 ODE
𝑑2𝑦 𝑑𝑦 nonlinear
equations for − 2𝑥 2 + 2𝑦 = 0
𝑑𝑥 2 𝑑𝑥 order 2
everything on
𝜕2 𝑁 PDE
this page: 𝜕𝑁
=
1 𝜕𝑁
+ r 𝑑𝑟 + 𝑘𝑁 nonlinear
𝑑𝑡 𝜕𝑟2
order 2
 Important DE Things Euler’s Method
Euler has this method to approximate f(c) using a point
on f(x) and f’(x).
Initial Value Problem (IVP)
It used to be done in Calc 1 & 2, but in Diff EQ, we will be
This is just a fancy way of explaining how to fully solve a using a slightly different formula:
DE (instead of having a C, you can actually have the real or 𝒚𝒏+𝟏 = 𝒚𝒏 + 𝒉 · 𝒇 𝒙𝒏, 𝒚𝒏
𝒚𝟏 = 𝒚𝟎 + 𝒉 · 𝒇 𝒙𝟎 , 𝒚𝟎
number)
IVP problems will give you y 𝑥0 = 𝑦0, and you plug 𝑥0and 𝑦0 are your initial points (which you will be given)
𝑥0, 𝑦0 into your formula once you solve it to solve for C. and h is the step size that you need to use. You’ll be given
(same with y′ 𝑥0 if given a 2nd order problem, and so on) those in the problem. You may need to use this formula
Existence and Uniqueness Solution Theorem many times to get your desired result.
This one is kinda weird... it basically means that if a
function and its differential equation are continuous on an Logistic Equations
interval (𝑎 < 𝑥0 < 𝑏 and 𝑐 < 𝑦0 < 𝑑), then there exists a This is just a fancy way of doing separation of variables
unique solution in that interval for (𝑥0, 𝑦0). from Calc AB/BC/1/2, but actually really fancy.
𝑥 2 −𝑦2
ex: If you’re given the IVP 𝑦 ′ = ,𝑦 𝑥0 = 𝑦0 , to see if the theorem applies, find
𝑥 2 +𝑦2
4𝑥 2 𝑦 • You’ll be given the direction field and you’ll need
𝑓 𝑥, 𝑦 , which I’ll just give to you: 𝑓 𝑥, 𝑦 = −
𝑥 2 +𝑦2 2
to predict some things.
Because 𝑓 𝑥, 𝑦 is continuous everywhere except 0, 0 , the E&U theorem implies that if
𝑥0, 𝑦0 ≠ 0, 0 , then the equation has a unique solution on some interval that contains 𝑥0, 𝑦0 • Just plug in the point given in the problem and
follow the trend to see where it ends up.
Direction Fields
𝑑𝑝
This is just a fancy way of saying slope fields! You should ex: You’re given the direction field for the DE =𝑝 𝑝−1 2−𝑝 ,
𝑑𝑡
remember how to do this from Calc AB/BC/1/2, but if you which is shown below: (p is in thousands)
forgot: What is lim 𝑝 𝑡 if 𝑝 0 = 4?
𝑡→∞
• They are basically, graph of the slope at every It is 2 (see the orange line)
point of a function
𝑑𝑦 What is lim 𝑝 𝑡 if 𝑝 0 = 1.7?
• They are also you get when you plug in x & y into 𝑑𝑥 𝑡→∞
It is 2 (see the green line)
Rules to know:
• if there is only x (no y), the slope will be the same for all x’s What is lim 𝑝 𝑡 if 𝑝 0 = 0.4?
𝑡→∞
• if there is only y (no x), the slope will be the same for all y’s It is 0 (see the purple line)

𝑑𝑦 𝑑𝑦 𝑑𝑦 Can a population of 900 ever

= 2x = 2y =x+y increase to 1000?
𝑑𝑥 𝑑𝑥 𝑑𝑥
No because when p(0) = 9, it goes
down to 2, not up to 10.
 ALL Linear Equations:
read
ing this, ta
ke a break and g
o watc
h snl

Types of DEs
are clip
ou s on
you

If y
tub
e, t
rus
They all follow this general equation: tm
ey

ou
’ll
Exact Equations: 𝑑𝑦

lov
e it
They all follow this general equation:
𝑎1 𝑥 + 𝑎0 𝑥 𝑦 = 𝑔 𝑥
𝑑𝑥
𝑴 𝒙, 𝒚 𝒅𝒙 + 𝑵 𝒙, 𝒚 𝒅𝒚 = 𝟎 However, this is the one we will use
And this must be true: when solving it (𝑎1 𝑥 = 1)
𝒅𝒚 Separable Equations: Equations of the Form:
𝝏𝑴 𝝏𝑵
𝒙, 𝒚 = 𝒙, 𝒚 +𝑷 𝒙 𝒚=𝑸 𝒙 They all follow this They all follow this general equation:
𝝏𝒚 𝝏𝒙 𝒅𝒙
general equation: 𝒅𝒚
How to solve them: = 𝑮 𝒂𝒙 + 𝒃𝒚 + 𝒄
How to solve them: 𝒅𝒚 𝒅𝒙
1. Change it to this form: = 𝒈 𝒙 𝒑(𝒚)
1. Change it to this form: 𝒅𝒙
𝑀 𝑥, 𝑦 𝑑𝑥 + 𝑁 𝑥, 𝑦 𝑑𝑦 = 0 How to solve them:
𝑑𝑦
+𝑃 𝑥 𝑦 =𝑄 𝑥 How to solve them: 1. Change it to this form:
2. Find the original function 𝑓(𝑥, 𝑦) 𝑑𝑥
that makes this true:Use methods from Calc 3
(remember gradients) 1. Change it to this form: 𝑑𝑦
2. Find the integrating factor: = 𝐺 𝑎𝑥 + 𝑏𝑦 + 𝑐
𝜕𝑓 𝜕𝑓 𝑑𝑥
𝜕𝑥
= 𝑀 𝑥, 𝑦 & 𝜕𝑦
= 𝑁 𝑥, 𝑦 µ 𝑥 = 𝑒 ‫𝑥𝑑 𝑥 𝑃 ׬‬
ℎ(𝑦)𝑑𝑦 = 𝑔 𝑥 𝑑𝑥
1 2. Transform your equation using this:
where ℎ 𝑦 = 𝑝 𝑦
3. Set it equal to C 3. Plug into this formula: dz 𝑑𝑦
Yes, I know there’s another method taught by your professors, but I think this is easier 2. Integrate both sides: 𝑧 = 𝑎𝑥 + 𝑏𝑦 + 𝑐 and dx = 𝑎 + 𝑏 𝑑𝑥
‫ ׬‬µ 𝑥 · 𝑄 𝑥 𝑑𝑥 + 𝐶
Integrating Factors: ) Not-Quite Exact
Equ ations ) 𝑦= 𝐻 𝑦 =𝐺 𝑥 +𝐶 3. Solve (hint: it’s separable)
µ(𝑥)
They all follow this general equation:
4. Change your variables back to y
µ𝑴 𝒙, 𝒚 𝒅𝒙 + µ𝑵 𝒙, 𝒚 𝒅𝒚 = 𝟎
Bernoulli Equations: Homogeneous Equations: and x using the above formulas
They all follow this general equation:They all follow this general equation:
That means that µ is a fu nction multiplied by both M and N
that makes them exact.

And this must be true: 𝒅𝒚 𝒅𝒚 𝒚

=𝑮
𝝏µ𝑴 𝝏µ𝑵 𝒅𝒙
+ 𝑷 𝒙 𝒚 = 𝑸 𝒙 𝒚𝒏 𝒅𝒙 𝒙 Important Info:
𝒙, 𝒚 = 𝒙, 𝒚 All magenta types are
𝝏𝒚 𝝏𝒙 How to solve them: How to solve them: solved like an exact
How to solve them: 1. Change it to this form: equation
1. Change it to this form: All blue types are solved
1. Change it to this form: 𝑑𝑦 𝑑𝑦 𝑦
+ 𝑃 𝑥 𝑦 = 𝑄 𝑥 𝑦𝑛 =𝐺 like a linear equation
𝑀 𝑥, 𝑦 𝑑𝑥 + 𝑁 𝑥, 𝑦 𝑑𝑦 = 0 𝑑𝑥 𝑥
𝜕𝑀 𝜕𝑁
𝑑𝑥 All orange types are solved
− 2. Transform your equation using these: like a separable equation
2. If 𝜕𝑦 𝜕𝑥
only depends on x, 𝜕𝑀 𝜕𝑁
− 2. Transform your equation into this: 𝑦 There is an example for
𝑁 𝜕𝑦 𝜕𝑥
𝝏𝑵 𝝏𝑴 µ 𝑥 = 𝑒 ‫𝑁 ׬‬ 𝑑𝑥
𝑑𝑣 𝑣 = and dy = vdx + xdv each of the 3 main types on
− 𝑥
3. If 𝝏𝒙 𝝏𝒚
only depends on y, 𝜕𝑁 𝜕𝑀
− + (1 − 𝑛)𝑃 𝑥 𝑣 = (1 − 𝑛)𝑄 𝑥 the next slide
𝑴 𝜕𝑥 𝜕𝑦
‫𝑦𝑑 𝑀 ׬‬
𝑑𝑥 3. Solve (hint: it’s separable)
µ 𝑦 = 𝑒 𝑣 = 𝑦1−𝑛
4. Change your variables back to y If youfollow my insta
4. Now use the steps for exact equations 3. Solve (hint: it’s linear) and x using the above formulas
to solve it, using your new equation: @JackYoung_71, I’ll send
4. Change your variables back to y you reels
µ𝑀 𝑥, 𝑦 𝑑𝑥 + µ𝑁 𝑥, 𝑦 𝑑𝑦 = 0 and x using the above formulas
The Main 3 Types of DEs (And Examples)
Separable Equations: Linear Equations: Exact Equations:
They all follow this general equation: They all follow this general equation: They all follow this general equation:
𝒅𝒚 𝑑𝑦 𝑴 𝒙, 𝒚 𝒅𝒙 + 𝑵 𝒙, 𝒚 𝒅𝒚 = 𝟎
= 𝒈 𝒙 𝒑(𝒚) 𝑎1 𝑥 + 𝑎0 𝑥 𝑦 = 𝑔 𝑥
𝒅𝒙 𝑑𝑥
And this must be true:
However, this is the one we will use
when solving it (𝑎1 𝑥 = 1) 𝝏𝑴 𝝏𝑵
𝒙, 𝒚 = 𝒙, 𝒚
𝒅𝒚 𝝏𝒚 𝝏𝒙
+𝑷 𝒙 𝒚=𝑸 𝒙
𝒅𝒙

How to solve them: How to solve them: How to solve them:

1. Change it to this form: 1. Change it to this form: 1. Change it to this form:
ℎ(𝑦)𝑑𝑦 = 𝑔 𝑥 𝑑𝑥 𝑑𝑦 𝑀 𝑥, 𝑦 𝑑𝑥 + 𝑁 𝑥, 𝑦 𝑑𝑦 = 0
+𝑃 𝑥 𝑦 =𝑄 𝑥
where ℎ 𝑦 =
1 𝑑𝑥 2. Find the original function
𝑝𝑦
2. Find the integrating factor: 𝑓(𝑥, 𝑦) that makes this true:
2. Integrate both sides: Use methods fr om Calc 3 (remember gradients)
µ 𝑥 = 𝑒‫𝑃 ׬‬ 𝑥 𝑑𝑥
𝜕𝑓 𝜕𝑓
𝐻 𝑦 =𝐺 𝑥 +𝐶
3. Plug into this formula: 𝜕𝑥
= 𝑀 𝑥, 𝑦 & 𝜕𝑦
= 𝑁 𝑥, 𝑦
3. Set it equal to C
‫ ׬‬µ 𝑥 · 𝑄 𝑥 𝑑𝑥 + 𝐶 Yes, I know there’s a sup er special method
𝑦= taught by your p rofessors, but I think th is is easier
µ(𝑥)

ex: ex: ex:

𝑑𝑦
1 + 𝑥 𝑑𝑦 − 𝑦𝑑𝑥 = 0 (𝑥 2
+ 9) + 𝑥𝑦 = 0 2𝑥 − 5𝑦 𝑑𝑥 + −5𝑥 + 3𝑦 2 𝑑𝑦 = 0
𝑑𝑥
𝑑𝑦 𝑥
1 1 + 2 𝑦=0
‫׬‬ 𝑑𝑦 = ‫׬‬ 𝑑𝑥 𝑑𝑥 𝑥 + 9 න2𝑥 − 5𝑦 𝑑𝑥 = 𝑥 2 − 5𝑦𝑥
𝑦 1+𝑥
𝑥
𝑃 𝑥 = 2 𝑄 𝑥 =0
ln 𝑦 = ln 1 + 𝑥 + 𝐶 𝑥 +9 න−5𝑥 + 3𝑦 2 𝑑𝑦 = −5𝑦𝑥 + 𝑦 3
𝑥
𝑒 ln 𝑦 = 𝑒 ln 1+𝑥 +𝐶 ‫ ׬‬2 𝑑𝑥
µ 𝑥 =𝑒 𝑥 +9 = 𝑥2 + 9 𝑓 𝑥, 𝑦 = 𝑥 2 − 5𝑦𝑥 + 𝑦 3 = 𝐶
|𝑦| = 𝑒 𝐶 |1 + 𝑥|
‫ 𝑥 ׬‬2 + 9 · 0𝑑𝑥 + 𝐶 𝐶 I p ut the two results of the integrals together to make
𝑦= = the function . See ho w this function becomes N and
y = ±𝐶(1 + 𝑥) 𝑥2 + 9 𝑥2 + 9 M when you take the partial derivativ e o f it
 Mass-Spring Oscillator
Applications You’re given the DE below that tells you how an oscillator
moves. It acts like a normal DE, the only thing is that you’ll
Orthogonal Trajectories have to solve for variables at the end.
They are essentially curves that intersects other curves at 1. You’ll be given an equation that will usually look something like this:
right angles when graphed. The definition of them doesn’t 𝒚 𝒕 = 𝑨𝒄𝒐𝒔 𝛀𝐭 + 𝐁𝐬𝐢𝐧(𝛀𝐭) Ω is just a random variable
matter as much as how to get them, which is below: 2. Calculate 𝑦′ and 𝑦′′
1. You’ll be given an equation with x, y, and k 3. Plug into this formula:𝒎𝒚′′ + 𝒃𝒚′ + 𝐤𝐲′ = 𝑭𝒆𝒙𝒕 𝐹𝑒𝑥𝑡 will be given in the question

Solve for 𝑓 𝑥, 𝑦 = 𝑘 4. Simplify and then solve for A and B

2. Calculate
𝜕𝑓
and
𝜕𝑓 Do this by setting up a system of equations by making
𝜕𝑥 𝜕𝑦 𝑐𝑜𝑠 Ωt = 0 for one equation and 𝑠𝑖𝑛 Ωt = 0 in another.
𝜕𝑓 𝜕𝑓
3. Plug into this formula: 𝜕𝑥 𝑑𝑦 = 𝜕𝑦 𝑑𝑥 5. Plug A and B back into your original 𝑦 𝑡
If you are ever asked if something oscillates, it does if it has cos(Ω t) and sin(Ωt) in it
4. Solve the DE (it’s separable)
Linearly Independent & Dependent Functions
Population Growth & Decay
A pair of functions are linearly dependent on an interval 𝐼 if
This is just a general set of equations that can describe the one of them multiplied by a constant is the other.
rate of change of a population of organisms with respect to If they can’t be, they are linearly independent.
time in proportion to the original population ex: 𝑦1 = csc 2 𝑡 − cot2 𝑡 𝑦2 = 7
𝑑𝑃 Sometimes you will be given a
= 𝑘𝑃 𝑃 = 𝐶 · ekt different DE in the question, that’s
𝑑𝑡 fine just solve it for P like usual 𝑦1 = 1... meaning 𝑦1 · 7 = 7 = 𝑦2
Law Cooling Formula:
7 is a constant, so they are
This is just a general set of equations that can describe the linearly dependent.
rate of change of a temp of something with respect to time
in proportion to the original temp and surrounding temp
𝑑𝑃
= 𝑘(𝑦 − 𝑠) 𝑦 = 𝐶 · ekt + s
𝑑𝑡
Variable Meanings For the Formulas Above: Things that are not math and are instead mostly physics
P = population C = initial amount (t = 0) I’m gonna be completely honest, there is only
one thing out of all the physics stuff that was
Y = temperature k is proportional constant tested on, which is resonance. Just know what
s = surrounding temperature k > 0 -> growth things cause resonance (like AM radio,
t = time (usually in years) k < 0 -> decay microwaves, opera singers, etc). Everything else
doesn’t matter.
Yes, this is from Calc BC/2 so you probably have seen this tl;dr- This is me if you ask me to explain physics:
before! spring stuff | resonance | population growth | microwave ovens | cooling law | AM radio | opera singers breaking glass | rota tional mechanics | series RLC circu
Homogeneous & Nonhomogeneous DEs
Homogeneous Linear Equations (When 𝒇(𝒕) = 𝟎) Nonhomogeneous Linear Equations (When 𝒇 𝒕 ≠ 𝟎)
This is just another type of DE that is easy to solve. This is just another type of DE that is stupid to solve.
They are of the form: They also are of the form:
𝒂𝒚′′ + 𝒃𝒚′ + 𝒄𝒚 = 𝒇 𝒕 = 𝟎 NOTICE THE 𝒇(𝒕) = 𝟎 𝒂𝒚′′ + 𝒃𝒚′ + 𝒄𝒚 = 𝒇 𝒕 ≠ 𝟎 NOTICE THE 𝒇 𝒕 ≠ 𝟎

Here is how you solve them (the homogeneous solution): Here is how you solve them (the particular solution):
1. You’ll be given the formula : 1. You’ll be given the formula :
𝑎𝑦 ′′ + 𝑏𝑦 ′ + 𝑐𝑦 = 𝑓 𝑡 𝑎𝑦 ′′ + 𝑏𝑦 ′ + 𝑐𝑦 = 𝑓 𝑡
2. Make the equation 𝒂𝒓𝟐 + 𝒃𝒓 + 𝒄 = 𝟎 2. Find what y 𝑡 is using the amazing table below
3. Solve for r (by quadratic formula or factoring)
3a. If you get two real solutions, your final answer will look like
𝒇(𝒕) 𝒚(𝒕)
this:𝒚 𝒕 = 𝒄𝟏 𝒆𝒓𝟏 𝒕 + 𝒄𝟐 𝒆𝒓𝟐 𝒕 𝒓 = 𝒓𝟏 , 𝒓𝟐 𝐶𝑡𝑚 𝐴𝑚 𝑡 𝑚 + ⋯ + 𝐴1 𝑡 + 𝐴0
3b. If you get one solution, your final answer will look like 𝐶𝑒 𝛼𝑡 𝐴𝑒 𝛼𝑡
this:𝒚 𝒕 = 𝒄𝟏 𝒆𝒓𝒕 + 𝒄𝟐 𝒕𝒆𝒓𝒕 𝒓 = 𝒓𝟏 = 𝒓𝟐
𝐶𝑠𝑖𝑛 𝛽𝑡 or Ccos 𝛽𝑡 𝐴𝑐𝑜𝑠 𝛽𝑡 + 𝐵𝑠𝑖𝑛 𝛽𝑡
3c. If you get two imaginary solutions, your final answer will look like
ex. if 𝑓 𝑡 = 𝐶𝑒 𝛼t sin 𝛽t , You do n’t need the C and A & B con stants, yo u
this:𝒚 𝒕 = 𝒄𝟏 𝒆𝜶𝒕 𝒄𝒐𝒔(𝜷𝒕) + 𝒄𝟐 𝒆𝜶𝒕 𝒔𝒊𝒏(𝜷𝒕) 𝒓 = 𝜶 ± 𝒊𝜷 Note: These for mulas wor k
when multiplied by each other then y t = 𝑒 𝛼𝑡 · (𝐴𝑐𝑜𝑠 𝛽𝑡 + 𝐵𝑠𝑖𝑛 𝛽𝑡 ) can just ch oose on e. Since A & B can be different,
choose those o ver the one C
The two c’s are just two different constants -> You’ll only get actual c values if the problem is an IVT
3. Solve your original formula like a homogenous equation and get
4. If it’s not an IVT, you’re done! If it is, plug in known values to your two solutions (r)
get a system of equations and then use those to find your 𝑐1 & 𝑐2
3a. If you have an equation where 𝑓 𝑡 = 𝐶𝑡 𝑚 , see if 0 is a
solution. Multiply your solution by 𝑡 𝑠 , using the info below for s:
Example: If 0 is not a solution, s = 0 If 0 is only one solution, s = If 0 is both solutions, s = 2
1
Homogeneous: Nonhomogeneous: 3b. If you have an equation where 𝑓 𝑡 = 𝐶𝑒 or 𝑓 𝑡 = 𝐶𝑡 𝑚 𝑒 𝛼𝑡 , see if 𝛼 is
𝛼𝑡
3
4𝑦 ′′ − 12𝑦 ′ + 9𝑦 = 0 4𝑦 ′′ − 12𝑦 ′ + 9𝑦 = 5𝑒 2
𝑡 a solution. Multiply your solution by 𝑡 𝑠 , using the info below for s:
If 𝛼 is not a solution, s = 0 If 𝛼 is only one solution, s = If 𝛼 is both solutions, s = 2
4𝑟 2 − 12𝑟 + 9 = 0 4𝑟 2 − 12𝑟 + 9 = 0 1
2 3c. If you have an equation where 𝑓 𝑡 = 𝐶𝑒 𝑠𝑖𝑛 𝛽t or 𝑓 𝑡 = 𝐶𝑒 𝛼𝑡 𝑐𝑜𝑠 𝛽t , see
𝛼𝑡
2𝑟 − 3 =0 2𝑟 − 3 2 =0 if 𝛼 + 𝑖𝛽 is a solution. Multiply your solution by 𝑡 𝑠 , using the info below for s:
3 3
3
𝑟 = → only one solution, so 𝑟 = → it’s both solutions to 5𝑒 2𝑡 , so If 𝛼 + 𝑖𝛽 is not a solution, s = 0 If 𝛼 + 𝑖𝛽 is a solution, s = 1
2 2
3
4. If you need to solve for your missing variables (A, 𝐴0, 𝐴1, B, C, etc.), find y 𝑡 ,
3 3
𝐴𝑡 2 𝑒 2𝑡 y′ 𝑡 , and y′′ 𝑡 and then plug them into your original equation. Now solve.
𝑦 𝑡 = 𝑐1 𝑒 2𝑡 + 𝑐1 𝑡𝑒 2𝑡 𝑦 𝑡 =
this is the part that makes it actual hell so good luck
You could solve for the You could solve for the Sometimes, you may be asked to do this but just solve for what form the equation is going to
constants if given more constants right now, I just don’t be in. For this, do not do step 4 and don’t plug in any variables that you already know (like 𝛼
information want to
or β)

these steps are extremely important for the following lessons so pay attention!
More Homogeneous & Nonhomogeneous Things
Cramer’s Rule was not tested on or in the hw, so dw about it!
Superposition Principle
Sometimes this takes a while, but it is very easy
If you have identical DEs with different f(t)’s, Variation of Parameters to remember the formulas. That’s why I love it.
you can just add their solutions This is a huge and great general method for finding a
If 𝐲𝟏 is a solution to 𝒂𝒚′′
+ 𝒃𝒚′
+ 𝒄𝒚 = 𝒇𝟏 𝒕 particular solution to a DE (I love it)
and
If 𝐲𝟐 is a solution to 𝒂𝒚′′ + 𝒃𝒚′ + 𝒄𝒚 = 𝒇𝟐 𝒕 1. You’ll be given an equation 𝑎𝑦 ′′ + 𝑏𝑦 ′ + 𝑐𝑦 = 𝑓 𝑡
2. Calculate the homogeneous solution, which will be c1y1 + c2y2
then...
3. Calculate 𝑊, 𝑊1, & 𝑊2 using the below formulas:
𝐜𝟏 𝐲𝟏 + 𝒄𝟐 𝒚𝟐 is a solution to 𝒂𝒚′′ + 𝒃𝒚′ + 𝒄𝒚 = 𝒄𝟏𝒇𝟏 𝒕 + 𝒄𝟐𝒇𝟐 𝒕
𝑦1 𝑦2 0 𝑦2 𝑦1 0 I d on’t un derstan d matrixes
So, if you ever have to solve a DE where f 𝑡 =__ + __ , you can just 𝑊 = 𝑦′
1 𝑦2′ 𝑊1 =
𝑓(𝑡) 𝑦2′
𝑊2 =
𝑦1′ 𝑓(𝑡)
either , just kno w you solve
them thro ugh this formula:
do each separately and add them later. Also, the 𝒄𝟏 and 𝒄𝟐 are
𝑊 = 𝑦1 𝑦2′ − 𝑦2 𝑦1′ 𝑊1 = −𝑦2 𝑓(𝑡) 𝑊2 = 𝑦1 𝑓(𝑡) 𝑎 𝑏
constants that let you use this formula, even if your 𝑦 𝑡 𝑠 and 𝐷=
𝑐 𝑑
= 𝑎𝑑 − 𝑏𝑐

𝑓 𝑡 s have been multiplied by a constant.

4. Solve for 𝑣1, & 𝑣2 using the below formulas:
General Solutions 𝑣1 = න
𝑊1
𝑑𝑡 𝑣2 = න
𝑊2
𝑑𝑡
𝑊 𝑊
𝑦𝑔 𝑡 = 𝑦ℎ (𝑡) + 𝑦𝑝 (𝑡) 5. Plug everything in to get your general solution (𝑦𝑝 )
𝒚𝒈 𝒕 is the the general solution to the DE 𝑎𝑦 + 𝑏𝑦 + 𝑐𝑦 = 𝑓 𝑡
′′ ′
𝑦𝑔 = 𝑦ℎ + 𝑦𝑝 = 𝑐1 𝑦1 + 𝑐2 𝑦2 + 𝑣1 𝑦1 + 𝑣2 𝑦2
𝒚𝒉 (𝒕) is the homogeneous solution, which is the specific solution to
𝑎𝑦 ′′ + 𝑏𝑦 ′ + 𝑐𝑦 = 0 (just set it equal to 0 to solve), without your constants Existence and Uniqueness Solution Theorem
solved for this is the easy part from a few lessons ago
This is very similar to the one before. Let’s say you have a
𝒚𝒑 (𝒕) is the particular solution (𝑦𝑝 (𝑡)), which is the specific solution to
𝑎𝑦 ′′ + 𝑏𝑦 ′ + 𝑐𝑦 = 𝑓 𝑡 , with all your constants solved for this is the bad p art DE in the following form: 𝒚′′ + 𝒑 𝒕 𝒚′ + 𝒒 𝒕 𝒚 = 𝒈(𝒕)f
from last lesso n

So, to solve these problems, solve for your particular If the functions 𝑝 𝑡 , 𝑞(𝑡) and 𝑔(𝑡) are continuous on an
solution first (including all variables). Then solve for your interval 𝑎 < 𝑡0 < 𝑏 then for any choice of the initial values
homogeneous solution first (by setting your DE equal to (𝑌0 & 𝑌1), there exists a unique solution for y 𝑡 in that
0). Finally add them together to create your general interval.
solution to the DE. You solve questions for them as follows:
Sometimes the problem might just give you what the 1. You’ll be given an equation 𝑓 𝑡 𝑦 ′′ + 𝑝 𝑡 𝑦 ′ + 𝑞 𝑡 𝑦𝑔(𝑡)
particular solution is, and if it does, you do not need to 2. Turn the equation into the form𝑦 ′′ + 𝑝 𝑡 𝑦 ′ + 𝑞 𝑡 𝑦 = 𝑔(𝑡) by
solve for it, you only need to solve for the homogeneous dividing the whole equation by 𝑓 𝑡 (You don’t need to do anything if 𝑓(𝑡) = 1)
solution and add them together. 3. Find the interval where all three of the functions (𝑝 𝑡 , 𝑞(𝑡) and 𝑔(𝑡) (in the
new form of the equation)) are continuous
4. Congrats! That’s your interview where there is a solution!
Cauchy-Euler Equations & Theorems
Cauchy-Euler Equations Reduction of Order Theorem
This is just a specific type of equation in the following If you are given one solution 𝑦1 𝑡 to the DE
form: 𝒅𝒏𝒚 𝒅𝒏−𝟏𝒚 𝒅𝒚 𝑦 ′′ + 𝑝 𝑡 𝑦 ′ + 𝑞 𝑡 𝑦 = 0, then you can use the
𝒏 𝒏−𝟏
𝒂𝒏 𝒕 + 𝒂𝒏−𝟏 𝒕 + ⋯ + 𝒂 𝟏 𝒕 + 𝒂𝟎𝒚 = 𝒇(𝒕) formula below to solve for the other solution 𝑦2 𝑡
𝒅𝒕𝒏 𝒅𝒕𝒏 𝒅𝒕
The degree of each term matches the order of the derivative.
Sometimes you might have to manipulate the equation to get it into this form 𝒆− ‫𝒕𝒅 𝒕 𝒑 ׬‬
𝒚𝟐 𝒕 = 𝒚𝟏 𝒕 න 𝒅𝒕
Ex: 3𝑡𝑦 ′′ + 5𝑦 ′ = 𝑡 4 is a Cauchy-Euler equation because if you multiply the 𝒚𝟐𝟏 𝒕
whole equation by 1, you get 3𝑡 2𝑦 ′′ + 5𝑡𝑦 ′ = 𝑡 5, which fits that formula.

Homogeneous Cauchy-Euler Equations

These are just Cauchy-Euler equations in the form of:
𝒂𝒕𝟐 𝒚′′ + 𝒃𝒕𝒚′ + 𝒄𝒚 = 𝟎
You solve them similarly to the normal homogeneous equations:
1. You’ll be given the formula : 𝑎𝑡 2 𝑦′′ + 𝑏𝑡𝑦′ + 𝑐𝑦 = 0
2. Make the equation 𝒂𝒓𝟐 + (𝒃 − 𝒂)𝒓 + 𝒄 = 𝟎
3. Solve for r (by quadratic formula or factoring) just have some empty space
3a. If you get two real solutions, your final answer will look like this:
𝒚𝒉 𝒕 = 𝒄𝟏 𝒕𝒓𝟏 + 𝒄𝟐 𝒕𝒓𝟐 𝒓 = 𝒓𝟏 , 𝒓𝟐
here, thought I’d just it to let you
3b. If you get one solution, your final answer will look like know I know these lessons are
this:𝒚𝒉 𝒕 = 𝒄𝟏 𝒕𝒓 + 𝒄𝟐 𝒕𝒓 𝒍𝒏 𝒕 𝒓 = 𝒓𝟏 = 𝒓𝟐 hard but you got thissss
3c. If you get two imaginary solutions, your final answer will look like this:
𝒚𝒉 𝒕 = 𝒄𝟏 𝒕𝜶 𝒄𝒐𝒔(𝜷𝒍𝒏 𝒕 ) + 𝒄𝟐 𝒕𝜶 𝒔𝒊𝒏(𝜷𝒍𝒏 𝒕 ) 𝒓 = 𝜶 ± 𝒊𝜷
The two c’s are just two different constants -> You’ll only get actual c values if the problem is an IVT

4. If it’s not an IVT, you’re done! If it is, plug in known values to

get a system of equations and then use those to find your 𝑐1 & 𝑐2

Non-Homogeneous Cauchy-Euler Equations

These are just Cauchy-Euler equations in the form of:
𝒂𝒕𝟐𝒚′′ + 𝒃𝒕𝒚′ + 𝒄𝒚 = 𝒇(𝒕)
You solve them using variation of parameters (see previous page):
Make sure when you solve the homogeneous version, you use the above method^^
 The formulas in bold are given to you on the exam (or a similar variation)

Laplace Transformations 𝓛{𝒇 𝒕 }

𝓛 𝒄𝒇
𝑭 𝒔 = 𝓛 {𝒇 𝒕 }
𝒄𝓛 𝒇

𝓛 𝒇+𝒈 𝓛 𝒇 +𝓛 𝒈
Laplace Transform
𝟏
𝓛{𝟏} ,𝒔 > 𝟎
If 𝑓(𝑡) is a function on [0, ∞), then the Laplace transform is: 𝒔
∞ 𝓛{𝒕𝒏} (𝐧 = 𝟏, 𝟐 … )
𝒏!
,𝒔 > 𝟎
𝓛 {𝒇(𝒕)} = 𝑭 𝒔 = න 𝒆−𝒔𝒕𝒇 𝒕 𝒅𝒕 𝒔𝒏+𝟏
𝟎 𝟏
𝓛{𝒆𝜶𝒕} ,𝒔 > 𝜶
It converts the domain of the function from 𝑡 to 𝑠 𝒔−𝒂

Existence of the Transform 𝓛{𝐬𝐢𝐧 𝒃𝒕 }

𝒃
𝒔𝟐 + 𝒃𝟐
,𝒔 > 𝟎

Piecewise Continuous Functions: ex: 𝓛{𝐜𝐨𝐬 𝒃𝒕 }

𝒔
,𝒔 > 𝟎
𝒔𝟐 + 𝒃𝟐
Functions that are continuous at every point except for
when it has a jump discontinuity on an interval. 2𝑏𝑠
ℒ{t sin 𝑏𝑡 } 2,𝑠 >0
𝑠2 + 𝑏2
Continuous Piecewise Functions: This was not in t he not es, I just
thought I would clarify the difference
ex:
𝑠2 − 𝑏2
ℒ{t cos 𝑏𝑡 } ,𝑠 > 0
Functions that are continuous at every point on a finite interval. 𝑠2 + 𝑏2 2

Exponential Order Functions: ex: 𝑓(𝑡) = 999𝑒 999𝑡 𝓛{𝐬𝐢𝐧 𝒉 𝒕 }

𝒃
, 𝒔 > |𝒃|
𝒔𝟐 − 𝒃𝟐
Functions that can fulfil the formula: Basically, it always It is of exponential order because it is
𝒔
𝒇𝒕 ≤ 𝑴𝒆𝜶𝒕 ( 𝑇 and 𝑀 are
positive constants ) needs to be less than
the function 𝑀𝑒 𝛼𝑡
always going to be smaller than 𝑔(𝑡) =
𝑀𝑒 𝛼𝑡 , where 𝑀 = 1000 and α = 1000
𝓛{𝐜𝐨𝐬 𝒉 𝒕 }
𝒔𝟐 − 𝒃𝟐
, 𝒔 > |𝒃|
Literally just replace any s in the formula with 𝑠 − 𝑎
More Theorems! 𝓛{𝒆𝜶𝒕𝒇 𝒕 } (𝐧 = 𝟏, 𝟐 … ) 𝑭 𝒔−𝒂 =𝓛 𝒇 𝒕 ቚ = 𝓛{𝒇}(𝒔 − 𝒂)
𝒔→𝒔−𝒂
If 𝑓(𝑡) is piecewise continuous on [0, ∞] and of Any other theorems
and formulas have 𝑛!
exponential order 𝛼, then ℒ {𝑓(𝑡)} exists for 𝑠 > 𝛼 been shortened and
ℒ{𝑡 𝑛 𝑒 𝛼𝑡 }
𝑠−𝑎 𝑛+1 , 𝑠 >𝛼

If 𝑓(𝑡) is piecewise continuous on 0, ∞ , of exponential thrown in the 𝒔−𝒂

table to the right 𝓛{𝒆𝜶𝒕𝐜𝐨𝐬(𝒃𝒕)} ,𝒔 > 𝜶
order 𝛼, and 𝐹 𝑠 = ℒ 𝑓 𝑡 , then lim 𝐹 𝑠 = 0 𝒔 − 𝒂 𝟐 + 𝒃𝟐
𝑠→∞ You might have to use partial 𝒃
Inverse Laplace Transformations! fractions to evaluate them, and
if so, check page 5 on my notes
𝓛{𝒆𝜶𝒕𝐬𝐢𝐧(𝒃𝒕)}
𝒔 − 𝒂 𝟐 + 𝒃𝟐
,𝒔 > 𝜶

The inverse of it, is where you go from the right −𝟏 ℒ{𝑡 𝑓 𝑡 } −

𝑑𝐹
𝓛 {𝑭(𝒔)} = 𝒇 𝒕 𝑑𝑠
side of the table to the left (the opposite) 𝒅𝒏
𝓛 {𝒕𝒏𝒇 𝒕 } −𝟏 𝒏 𝓛{𝒇 𝒕 }
Solving IVTs All the formulas you need are to the right! 𝒅𝒔𝒏
𝓛{𝒇′ 𝒕 } 𝒔𝓛 𝒇 𝒕 − 𝒇(𝟎)
1. Take the Laplace transform of both sides of the equation
𝓛{𝒇′ ′ 𝒕 } 𝒔𝟐 𝓛 𝒇 𝒕 − 𝒔𝒇 𝟎 − 𝒇′ 𝟎
2. Solve for ℒ 𝑦 𝑠 (which is Y 𝑠 )
3. Take the inverse Laplace transform of both sides to solve for 𝑦 𝑠 𝓛 {𝒇(𝒏) 𝒕 } 𝒔𝒏𝓛 𝒇 𝒕 − 𝒔𝒏−𝟏𝒇 𝟎 − 𝒔𝒏−𝟐𝒇′ 𝟎 − ⋯ − 𝒇 𝒏−𝟏 (𝟎)

If your initial conditions are not at 𝑡 = 0, use the substitution 𝑦 𝑡 = 𝑤 𝑡 + 𝑐 , where 𝓛{𝒕𝒚′}(𝒔) −𝒔𝒀′ 𝒔 − 𝒀 𝒔 , where 𝒀 𝒔 = 𝓛{𝒚}(𝒔)
𝑤 is your given function and 𝑐 is the 𝑡 value where your initial conditions are given 𝓛{𝒕𝒚′′}(𝒔) −𝒔𝟐𝒀′ 𝒔 − 𝟐𝒔𝒀 𝒔 , +𝒚 𝟎 , where 𝒀 𝒔 = 𝓛{𝒚}(𝒔)
 Unit Step, Convolution, & Dirac Delta
Discontinuous Functions Transfer Function
The unit step function: Transformed unit step function: If you have an equation of the form
𝟎, 𝒕 < 𝟎 𝟎, 𝒕 < 𝒂 𝑀
𝑎𝑦 ′′ + 𝑏𝑦 ′ + 𝑐𝑦 = 𝑔 𝑡 ; 𝑦 0 = 𝑦 ′ 0 = 0
𝒖 𝒕 =ቊ 𝑴·𝒖 𝒕−𝒂 =ቊ You can use the following transfer function (𝐻(𝑠)) to study the
𝟏, 𝒕 > 𝟎 𝑴, 𝒕 > 𝒂
𝑎
behavior of the system for different inputs.
You can transform piecewise functions into unit
step functions using the following formula: 𝑌 𝑠 1 1
𝐻 𝑠 = = 2 ℎ 𝑡 = ℒ −1 𝐻(𝑠) = ℒ −1 2
𝐺(𝑠) 𝑎𝑠 + 𝑏𝑠 + 𝑐 𝑎𝑠 + 𝑏𝑠 + 𝑐
𝒇𝒕 , 𝒂<𝒕<𝒃
Impulse Response Function
if 𝒚 𝒕 = ൞𝒈 𝒕 , 𝒃 < 𝒕 < 𝒄, then
ℎ(𝑡) is called the impulse response function, and it is used for
𝒉𝒕 , 𝒄<𝒕
positive when 𝑡 is greater than a # mass-spring systems struck by a hammer. You can solve IVTs with
𝒚 𝒕 =𝒇 𝒕 −𝒇 𝒕 ·𝒖 𝒕−𝒃 +𝒈 𝒕 ·𝒖 𝒕−𝒃 it using the steps below:
−𝒈 𝒕 · 𝒖 𝒕 − 𝒄 + 𝒉(𝒕) · 𝒖 𝒕 − 𝒄 1. Let 𝐼 be an interval containing the origin and let 𝑔 be continuous on 𝐼.
negative when 𝑡 is less than a # You’ll be given: 𝑎𝑦 ′′ + 𝑏𝑦 ′ + 𝑐𝑦 = 𝑔 𝑡 ; 𝑦 0 = 𝑦𝑜 ; 𝑦 ′ 0 = 𝑦1
Laplace Transformations 2. Find the homogenous solution 𝑦ℎ (𝑡)
𝒆−𝒂𝒔 3. Find the impulse response function using the formula above:
𝓛 𝒖 𝒕−𝒂 = 𝒔
, for 𝑠 >0 𝑡
4. The solution is: y 𝑡 = 𝑦ℎ 𝑡 + h ∗ g = 𝑦ℎ 𝑡 + න ℎ 𝑡 − 𝑣 𝑔 𝑣 𝑑𝑣
𝓛 𝒇(𝒕 − 𝒂)𝒖 𝒕 − 𝒂 = 𝒆−𝒂𝒔𝓛{𝒇 𝒕 } for 𝑠 > 𝑎 ≥ 0 0

If 𝑔 𝑡 = 𝑓(𝑡 − 𝑎), then 𝑓 𝑡 = 𝑔 𝑡 + 𝑎 , so Dirac Delta Function It is 0 everywhere but is ∞ at 𝑡 = 0

∞
𝓛 𝒈(𝒕)𝒖 𝒕 − 𝒂 = 𝒆−𝒂𝒔𝓛{𝒈 𝒕 + 𝒂 } for 𝑠 > 𝑎 ≥ 0 𝟎, 𝒕≠𝟎
𝜹 𝒕 =ቊ න 𝒇(𝒕)𝜹 𝒕 𝒅𝒕 = 𝒇(𝟎)
∞, 𝒕=𝟎 −∞
Try and rewrite the functions to be 𝑓(𝑡 − 𝑎) so that you can use the
above formulas. It might be easier to do that and then simplify. Transformations & Formulas of it Include: These are true if the 0/𝑎 is
inside of limits of the integral
The opposite of these works for inverse Laplace transforms 𝟎, 𝒕 ≠ 𝒂 ∞
𝜹 𝒕−𝒂 =ቊ න 𝒇 𝒕 𝜹 𝒕 − 𝒂 𝒅𝒕 = 𝒇(𝒂) 𝓛 𝜹 𝒕−𝒂 = 𝒆−𝒂𝒔
Convolution ∞, 𝒕 = 𝒂 −∞
It is an operation to help solve Laplace transforms: 𝒕
𝒕 න 𝜹 𝒙 − 𝒂 𝒅𝒙 = 𝒖 𝒕 − 𝒂 𝜹 𝒕 − 𝒂 = 𝒖′ 𝒕 − 𝒂
ex. 𝑡 ∗ 1 𝑡
𝒇 ∗ 𝒈 𝒕 = න 𝒇 𝒕 − 𝒗 𝒈 𝒗 𝒅𝒗 𝒕 𝒕𝟐
−∞
= ‫ 𝒕 𝟎׬‬− 𝒗 𝟏 𝒅𝒗 =
Solving IVTs with the Dirac Delta Function
(𝑓 ∗ 𝑔) is n’ t m ultipl ied by 𝑡, thos e
functi ons are of 𝑡, lik e 𝑓(𝑡) 𝟎 𝟐

Properties & More Formulas

The IVT: 𝑎𝑦 ′′ + 𝑏𝑦 ′ + 𝑐𝑦 = 𝛿 𝑡 ; 𝑦 0 = 0; 𝑦 ′ 0 = 0 is solved like normal
𝑓∗0=0 ℒ 𝑓∗𝑔 =𝐹 𝑠 𝐺 𝑠 ℒ −1 𝐹 𝑠 𝐺 𝑠 =𝑓∗𝑔
(through laplace transforms), but you can use this formula to solve for y:
𝑓∗ 𝑔+ℎ = 𝑓∗𝑔 + 𝑓∗ℎ 𝑡 𝑡
𝐹 𝑠 𝐹 𝑠 You can also use the impulse response function
𝑓 ∗ 𝑔 ∗ ℎ = 𝑓 ∗ (𝑔 ∗ ℎ) ℒ න 𝑓 𝑣 𝑑𝑣 =
𝑠
ℒ −1
𝑠
= න 𝑓 𝑣 𝑑𝑣 ℒ −𝟏{𝒆−𝒂𝒔 · 𝑭(𝒔)} = 𝒖(𝒕 − 𝒂) · 𝒇(𝒕) to solve it when the situation calls for it.
𝑓∗𝑔=𝑔∗𝑓 0 0
 The formulas in bold are given to you on the exam (or a similar variation)
Just a page with every important formula for this unit (so far)

Laplace Formulas 𝓛{𝒇 𝒕 }

𝓛 𝒄𝒇
𝑭 𝒔 = 𝓛 {𝒇 𝒕 }
𝒄𝓛 𝒇

Laplace ∞
𝓛 𝒇+𝒈 𝓛 𝒇 +𝓛 𝒈

ℒ {𝑓(𝑡)} = 𝐹 𝑠 = න 𝑒 −𝑠𝑡 𝑓 𝑡 𝑑𝑡 ℒ −1 {𝐹(𝑠)} = 𝑓 𝑡 𝓛{𝟏}

𝟏
,𝒔 > 𝟎
0 𝒔
Unit Step Function 𝓛{𝒕𝒏} (𝐧 = 𝟏, 𝟐 … )
𝒏!
,𝒔 > 𝟎
𝒔𝒏+𝟏
0, 𝑡<0 0, 𝑡<𝑎
𝑢 𝑡 =ቊ 𝑀·𝑢 𝑡−𝑎 =ቊ 𝟏
1, 𝑡>0 𝑀, 𝑡>𝑎 𝓛{𝒆𝜶𝒕} ,𝒔 > 𝜶
𝒔−𝒂
𝒃
𝑓 𝑡 , 𝑎<𝑡<𝑏 𝓛{𝐬𝐢𝐧 𝒃𝒕 }
𝒔𝟐 + 𝒃𝟐
,𝒔 > 𝟎
𝑓 𝑡 −𝑓 𝑡 ·𝑢 𝑡−𝑏 +𝑔 𝑡 ·𝑢 𝑡−𝑏
𝑦 𝑡 = ൞𝑔 𝑡 , 𝑏 < 𝑡 < 𝑐 = 𝒔
− 𝑔 𝑡 · 𝑢 𝑡 − 𝑐 + ℎ(𝑡) · 𝑢 𝑡 − 𝑐 𝓛{𝐜𝐨𝐬 𝒃𝒕 } ,𝒔 > 𝟎
ℎ 𝑡 , 𝑐<𝑡 𝒔𝟐 + 𝒃𝟐
𝑒 −𝑎𝑠 2𝑏𝑠
ℒ 𝑢 𝑡−𝑎 = , for 𝑠>0 ℒ 𝑓(𝑡 − 𝑎)𝑢 𝑡 − 𝑎 = 𝑒 −𝑎𝑠 ℒ{𝑓 𝑡 } ℒ{t sin 𝑏𝑡 } 2,𝑠 >0
𝑠 𝑠2 + 𝑏2
If 𝑔 𝑡 = 𝑓(𝑡 − 𝑎), then 𝑓 𝑡 = 𝑔 𝑡 + 𝑎 , so ℒ 𝑔(𝑡)𝑢 𝑡 − 𝑎 = 𝑒 −𝑎𝑠 ℒ{𝑔 𝑡 + 𝑎 } 𝑠2 − 𝑏2
ℒ{t cos 𝑏𝑡 } ,𝑠 > 0
𝑠2 + 𝑏2 2
Convolution
𝒃
𝑡 𝓛{𝐬𝐢𝐧 𝒉 𝒕 } , 𝒔 > |𝒃|
𝒔𝟐 − 𝒃𝟐
𝑓 ∗ 𝑔 𝑡 = න 𝑓 𝑡 − 𝑣 𝑔 𝑣 𝑑𝑣
0 𝒔
𝓛{𝐜𝐨𝐬 𝒉 𝒕 } , 𝒔 > |𝒃|
ℒ −1 𝐹 𝑠 𝐺 𝑠 =𝑓∗𝑔 𝒔𝟐 − 𝒃𝟐
𝑓∗0=0 ℒ 𝑓∗𝑔 =𝐹 𝑠 𝐺 𝑠 Literally just replace any s in the formula with 𝑠 − 𝑎
𝑓∗𝑔 =𝑔∗𝑓 𝑡
𝐹 𝑠 𝐹 𝑠 𝑡
𝓛{𝒆𝜶𝒕𝒇 𝒕 } (𝐧 = 𝟏, 𝟐 … ) 𝑭 𝒔−𝒂 =𝓛 𝒇 𝒕 ቚ = 𝓛{𝒇}(𝒔 − 𝒂)
ℒ න 𝑓 𝑣 𝑑𝑣 = ℒ −1 = න 𝑓 𝑣 𝑑𝑣 𝒔→𝒔−𝒂
𝑓 ∗ 𝑔 ∗ ℎ = 𝑓 ∗ (𝑔 ∗ ℎ) 0 𝑠 𝑠 0
𝑛!
𝑓∗ 𝑔+ℎ = 𝑓∗𝑔 + 𝑓∗ℎ ℒ{𝑡 𝑛 𝑒 𝛼𝑡 } 𝑛+1 , 𝑠 >𝛼
𝑠−𝑎
Transfer & Impulse Response Functions 𝒔−𝒂
𝓛{𝒆𝜶𝒕𝐜𝐨𝐬(𝒃𝒕)} ,𝒔 > 𝜶
𝒔 − 𝒂 𝟐 + 𝒃𝟐
𝑌 𝑠 1
𝐻 𝑠 = = 2 ℎ 𝑡 = ℒ −1 𝐻(𝑠) 𝒃
𝐺(𝑠) 𝑎𝑠 + 𝑏𝑠 + 𝑐 𝓛{𝒆𝜶𝒕𝐬𝐢𝐧(𝒃𝒕)} ,𝒔 > 𝜶
𝒔 − 𝒂 𝟐 + 𝒃𝟐
Dirac Delta Function ℒ{𝑡 𝑓 𝑡 } −
𝑑𝐹
∞ 𝑑𝑠
0, 𝑡≠0
𝛿 𝑡 =ቊ න 𝑓(𝑡)𝛿 𝑡 𝑑𝑡 = 𝑓(0) 𝓛 {𝒕𝒏𝒇 𝒕 }
𝒅𝒏
∞, 𝑡=0 −∞ −𝟏 𝒏
𝒅𝒔𝒏
𝓛{𝒇 𝒕 }
∞
0, 𝑡≠𝑎 𝓛{𝒇′ 𝒕 } 𝒔𝓛 𝒇 𝒕 − 𝒇(𝟎)
𝛿 𝑡−𝑎 =ቊ න 𝑓 𝑡 𝛿 𝑡 − 𝑎 𝑑𝑡 = 𝑓(𝑎)
∞, 𝑡=𝑎 −∞ 𝓛{𝒇′ ′ 𝒕 } 𝒔𝟐 𝓛 𝒇 𝒕 − 𝒔𝒇 𝟎 − 𝒇′ 𝟎
𝑡
𝛿 𝑡 − 𝑎 = 𝑢′ 𝑡 − 𝑎 𝓛 {𝒇(𝒏) 𝒕 } 𝒔𝒏𝓛 𝒇 𝒕 − 𝒔𝒏−𝟏𝒇 𝟎 − 𝒔𝒏−𝟐𝒇′ 𝟎 − ⋯ − 𝒇 𝒏−𝟏 (𝟎)
න 𝛿 𝑥 − 𝑎 𝑑𝑥 = 𝑢 𝑡 − 𝑎
−∞ 𝓛{𝒕𝒚′}(𝒔) −𝒔𝒀′ 𝒔 − 𝒀 𝒔 , where 𝒀 𝒔 = 𝓛{𝒚}(𝒔)

ℒ 𝛿 𝑡−𝑎 = 𝑒 −𝑎𝑠 ℒ −1 {𝑒 −𝑎𝑠 · 𝐹(𝑠)} = 𝑢(𝑡 − 𝑎) · 𝑓(𝑡) 𝓛{𝒕𝒚′′}(𝒔) −𝒔𝟐𝒀′ 𝒔 − 𝟐𝒔𝒀 𝒔 , +𝒚 𝟎 , where 𝒀 𝒔 = 𝓛{𝒚}(𝒔)
 The following lessons include series in
them.
I included my Calc 2/BC Series notes here
to hopefully help & give a review.
Series “When in doubt, ratio test” - Jack 2023
Types of Series
Review
All the Tests to See If a Series Converges or Diverges • Factorial:
∞
෍
1
𝑝
1 1 1
= 1+1+ + + +⋯ It converges
∞ 𝑛 2 6 24 𝑎
Limit Test 𝑛=1 If 𝑟 < 1, it converges to 1−𝑟
• Geometric: ෍ 𝑎𝑟 𝑛−1 = 𝑎 + 𝑎𝑟 + 𝑎𝑟2 + ⋯ , 𝑎 ≠ 0
If 𝑟 ≥ 1, it diverges
• Find the lim 𝑎𝑛 of the series. If lim 𝑎𝑛 ≠ 0, then the series diverges. 𝑛=1
∞
𝑛→∞ 𝑛→∞ 1 1 1 If p > 1, it converges
Integral Test • p-series: ෍ =1+ 𝑝+ 𝑝+⋯
∞ 𝑛𝑝 2 3 If 0 < 𝑝 ≤ 1, it diverges
𝑛=1
• If the series෍ 𝑎𝑛 only has positive terms and the integral is easy to find: ∞
𝑛=1
• Telescoping෍(𝑏𝑛 − 𝑏𝑛+1) Sum: 𝑆 = 𝑏`1 − 𝐿 If lim 𝑏𝑛 = 𝐿, it converges
• Set 𝑎𝑛 = 𝑓 𝑥 and make sure 𝑓 𝑥 is continuous, positive, and decreasing 𝑛→∞

∞ : 𝑛=1
∞
• If it is, solve ‫׬‬1 𝑓 𝑥 𝑑𝑥 ∞ ∞ • Harmonic: ෍ 1 = 1 + 1 + 1 + ⋯ It diverges
𝑛 2 3
• If the integral converges,෍ 𝑎𝑛 converges. If the integral diverges,
෍ 𝑎𝑛 diverges 𝑛=1
𝑛=1 𝑛=1 ∞
• k-to-the-k: 1 1 1
Limit Comparison Test ෍ =1+ 2+ 3+⋯ It converges
∞ 𝑛𝑛 2 3
𝑛=1
• If the series ෍ 𝑎𝑛 has only positive terms and is a fraction with a polynomial ∞
𝑛=1 • Alternating −1 𝑛+1 1 1 1
It converges
as the denominator, then Harmonic:
෍
𝑛
= 1− + − +⋯
2 3 4
𝑛=1
𝑎𝑛
• Make another sequence (𝑏𝑛 ) and find lim = 𝐿. If 𝐿 > 0, then

•
∞ ∞
𝑛→∞ 𝑏∞
𝑛

If ෍ 𝑏𝑛converges, then ෍ 𝑎𝑛converges. If ෍ 𝑏𝑛 diverges, then ෍ 𝑎𝑛diverges

∞ Power Series
∞

Looks like this: ෍ 𝑎𝑛 𝑥 − 𝑐 𝑛

𝑛=1 𝑛=1 𝑛=1 𝑛=1
Comparison Test∞
𝑛=0
• If the series෍ 𝑎𝑛 has only positive terms, then make another sequence (𝑏𝑛 )
𝑛=1 ∞ ∞ Each series has an interval of convergence, which includes all x values
• If 0 < 𝑎𝑛 ≤ 𝑏𝑛 and ෍ 𝑏𝑛converges, then ෍ 𝑎𝑛converges that make the series converge.
𝑛=1
∞ ∞
𝑛=1
Find the interval of convergence by doing ratio test and solving for x.
• If 0 < 𝑏𝑛 ≤ 𝑎𝑛 and ෍ 𝑏𝑛 diverges, then෍ 𝑎𝑛 diverges
Alternating Series Test
𝑛=1 𝑛=1 The midpoint/center of the interval of convergence is C, and the
•
∞

If the series෍ −1 𝑛−1 𝑎

is not increasing, 0 < 𝑎𝑛+1 ≤ 𝑎𝑛 , and lim 𝑎𝑛 = 0,
𝑛
radius of convergence is how far it is from C to either endpoint
𝑛=1 𝑛→∞
then it converges
Absolute Convergence Test
Ex:
∞
Find the radius of convergence for the series
∞
∞
𝑘𝑥 𝑘 𝑘 𝑐 = 0, so the center of the
• If the series෍ −1 𝑛−1 𝑎
converges in any other test but without the −1 𝑛−1 part, ෍ 𝑘 = ෍ 𝑘 (𝑥 − 0)𝑘
𝑛
4 4 radius of convergence is 0
𝑛=1 𝑘=0 𝑘=0
it is absolutely convergent. If it diverges, it might be conditionally convergent. Ratio test:
Ratio Test ∞ 𝑘 + 1 (𝑥 𝑘+1 ) 𝑥
𝑘 + 1 (𝑥 𝑘 )(𝑥) 4𝑘 𝑘𝑥 + 𝑥
• If the series෍ 𝑎𝑛 has nonzero terms that involve products, factorials, or powers, lim 4 𝑘+1 lim · 𝑘 lim lim <1 −4 < 𝑥 < 4
𝑘→∞ 𝑘𝑥 𝑘 𝑘→∞ (4𝑘 )(4) 𝑘𝑥 𝑘→∞ 4𝑘 𝑘→∞ 4
𝑛=1 4 You g et
∞
for the l im it, s o us e LHopital ’s Rul e
∞

𝑎𝑛+1
• Find lim . If it is < 1, it converges. If it is > 1 or ∞, it diverges. Test endpoints (-4 & 4):
𝑛→∞ 𝑎𝑛
Root Test
∞ ∞
∞ ෍
𝑘(4)𝑘
= ෍𝑘 Diverges The interval of convergence is -4 < x < 4
4𝑘
• If the series෍ 𝑎𝑛 has nonzero terms and the nth term involves an nth power 𝑘=0
∞
𝑘=0
∞
(4 not included in
int erval)
The center is 0, and because the distance
𝑛
𝑛=1 ෍
𝑘(−4)𝑘
4𝑘
= ෍ −1 𝑘 𝑘 Diverges from 0 to 4 is 4, the radius of convergence
• Find lim 𝑎𝑛 . If it is < 1, it converges. If it is > 1 or ∞, it diverges. 𝑘=0 𝑘=0
is 4
𝑛→∞
Taylor & Maclaurin Series (PAIN.) Review
Taylor Series Representing Functions as Maclaurin Series Formulas
f(x) =
∞
𝑓 𝑘 𝑐 MEMORIZE THESE
𝑓 𝑥 =෍ 𝑥−𝑐 𝑘
𝑘! f(x) =
𝑘=0
f(x) =
Maclaurin Series (c = 0)
f(x) =
∞ 𝑘
𝑓 0 𝑘 “When in doubt, skip
𝑓 𝑥 =෍ 𝑥 f(x) =
𝑘! these questions, you’ll
𝑘=0 probably pass
f(x) = anyways” - Jack 2023
Because it is a type of power series, it
also has an interval of convergence,
along with a radius of convergence. f(x) =

C is the center of the radius of convergence.

f(x) =
You can find the interval of
f(x) =
convergence using the ratio test
Look at the previous page for
the power series to see how This formula helps a ton for simplification:
(𝒙 + 𝟏)! = (𝒙 + 𝟏) · (𝒙)!
They are used to approximate
values of f(x) f(a), plugging in a for x
The more terms in the series you See the next slide for fun examples!
add, the closer the approximation They actually help for the diff eq lessons
will be.
Taylor & Maclaurin Series EXAMPLES Review
(PAIN.)
Ex: Deriving the Maclaurin series for f(x) = ex

fk fk(0)
k=0 f’(x) = ex f’(0) = 1
From the first few terms of the series, you can rewrite it to something simpler:
k=1 f(x) = ex f(0) = 1
The terms follow a pattern of:
k=2 f2(x) = ex f2(0) = 1
k=3 f3(x) = ex f3(0) = 1 The series made from these terms is:
k=4 f4(x) = ex f4(0) = 1
k=5 f5(x) = ex f5(0) = 1

Ex: If f(x) = ln(x), make a Maclaurin series for it and approximate f(2)
∞ ∞
𝑥 𝑘+1 (𝑥 − 1)𝑘+1
Since ln 1 + 𝑥 = ෍ −1 𝑘
𝑘+1
, ln 𝑥 = ln 1 + 𝑥 − 1 = ෍ −1 𝑘
𝑘+1
𝑘=0 𝑘=0

f(2) =

ln(2) = 0.693... “close enough”

 Shifting the Summation Index
NOOOOO SERIES :( Note: I shorten this to “sum index” for my notes
To rewrite a series with a different sum index, you can do the
following:
The Taylor Polynomial Approximation
1. Shift the sum index up by 𝑐 and all 𝑛’s down by 𝑐
Essentially, this lesson has nothing new. It combines the
2. Shift the sum index down by 𝑐 and all 𝑛’s up by 𝑐
basics of differential equations with the Taylor Series. If you ∞ ∞
forgot anything in those two topics, look at the slides before ex: ෍ 𝒏 𝒏 − 𝟗 𝒂𝒏 𝒙𝒏+𝟖 = ෍ (𝒏 − 𝟖) 𝒏 − 𝟏𝟕 𝒂𝒏−𝟏𝟕 𝒙𝒏
this. 𝒏=𝟖 𝒏=𝟏𝟔
To solve these types of questions, use what you know about Simple Operations on Power ∞Series
calc to find each term. I personally like using a table to solve ∞

If 𝒇 𝒙 = ෍ 𝒂𝒏 (𝒙 − 𝒙𝟎)𝒏 and 𝒈 𝒙 = ෍ 𝒃𝒏 (𝒙 − 𝒙𝟎)𝒏, then...

them.
𝒏=𝟎 (for all x in interval 𝐼1) 𝒏=𝟎 (for all x in interval 𝐼2)
Ex. Find the 4th degree Taylor Polynomial for 𝑦 ′′ + 𝑦 = 𝑒 2𝑥 about 𝑥0 = 0, ∞
given 𝑦(0) = 0 and 𝑦’(0) = 1 𝒇 𝒙 + 𝒈(𝒙) = ෍(𝒂𝒏 +𝒃𝒏 )(𝒙 − 𝒙𝟎)𝒏 for all x in 𝐼1 ∩ 𝐼2
∞ 𝒏=𝟎
𝒇(𝒏) 𝒙𝟎 𝒌 ∞
𝒇(𝒏) (𝒙) 𝒇(𝒏) (𝒙) 𝒇∗𝒏) 𝒙𝟎 ෍
𝒏!
𝒙 − 𝒙𝟎 ∞
You g et thes e by taki ng the derivativ e of the previous term
You g et thes e by plug g ing in 𝑥0
and all the 𝑦 val ues that you
𝒏=𝟎
You g et thes e by plug g ing in ev ery thi ng i nto that form ula above^ )𝒏
𝒇 𝒙 · 𝒈(𝒙) = ෍(𝒄𝒏 )(𝒙 − 𝒙𝟎 ,where 𝒄𝒏 = ෍ 𝒂𝒌𝒃𝒏−𝒌 for all x in 𝐼1 ∩ 𝐼2
have

𝑦 0 0 ex: 𝒏=𝟎
∞
𝒌=𝟎
∞
We’re given 𝒇𝒏(𝒙𝟎) 𝟑 ෍ 𝒂𝒏 𝒙𝒏+𝟏 + ෍ 𝒏𝒃𝒏 𝒙𝒏−𝟏
Simplify the following:
𝑦′ so these don’t matter 1 𝑥 𝒏=𝟏 𝒏=𝟐

∞ ∞
𝑦′′ 𝑦 ′′ = 𝑒 2𝑥 − 𝑦 1 𝑥2 Make the the two be in terms of 𝑥 𝑛
Solve for 𝑦’’ in the given equation
by shifting the sum index.
𝟑 ෍ 𝒂𝒏−𝟏𝒙𝒏 + ෍(𝒏 + 𝟏)𝒃𝒏+𝟏𝒙𝒏
1 3 𝒏=𝟐 𝒏=𝟏
𝑦′′′ 𝑦 ′′′ = 2𝑒 2𝑥 − 𝑦′ 1 𝑥
6 Make the the two have the same
∞ ∞

1 4 summation index by just adding 𝟑 ෍ 𝒂𝒏−𝟏𝒙𝒏 + 𝟐𝒃𝟐𝒙 + ෍(𝒏 + 𝟏)𝒃𝒏+𝟏𝒙𝒏

𝑦 (4) 𝑦 (4) = 4𝑒 2𝑥 − 𝑦′′ 3 𝑥 the terms before 𝑛 = 2 𝒏=𝟐 𝒏=𝟐
8
∞
1 1 Use the formula above to
𝑃𝑛 𝑥 = 0 + 𝑥 + 𝑥 2 + 𝑥 3 + 𝑥 4 𝟐𝒃𝟐𝒙 + ෍(𝒂𝒏−𝟏 + (𝒏 + 𝟏)𝒃𝒏+𝟏)𝒙𝒏
6 8 combine them
𝒏=𝟐
Old Power Series Stuff
Calculus Operations on Power Series
To find out how to write out power series, find the interval of ∞

convergence, etc, use the previous slides, it’s all there. If 𝒇 𝒙 = ෍ 𝒂𝒏 (𝒙 − 𝒙𝟎)𝒏 = 𝒂𝟎 + 𝒂𝟏 𝒙 − 𝒙𝟎 + 𝒂𝟐 𝒙 − 𝒙𝟎 𝟐 + ⋯ , then...
𝒏=𝟎 (for 𝑥 − 𝑥0 < 𝜌)
∞ ∞
Multiplying Power Series ෍ 𝒏𝒂𝒏 (𝒙 − 𝒙𝟎)𝒏−𝟏
𝒂𝒏
𝒇′ 𝒙 = and න𝒇 𝒙 = 𝑪 + ෍
𝒏+𝟏
(𝒙 − 𝒙𝟎)𝒏+𝟏
To find the first few terms of the power series for 𝑓 𝑥 · g(x), just 𝒏=𝟏 (for 𝑥 − 𝑥0 < 𝜌)
Notice the 𝒏 = 𝟏, when you take the
𝒏=𝟎 (for 𝑥 − 𝑥0 < 𝜌)

write out the terms for each and multiply them together. Write deriva tive, your sum index will go up by 1.

them in order of their power.

 Minimum Radius of Convergence
More Series Fun! In the default power series form, 𝑦 𝑥 = ෍ 𝑎𝑛 𝑥 − 𝑥0 𝑛 ,
𝑦 converges on the interval 𝑥 − 𝑥0 < 𝑅,𝑛=0 where 𝑅 is the
∞

Analytic Functions distance from 𝑥0. You calculate this by solving for all the
A function is analytic at 𝑥 = 𝑥0 if there is an open interval singular points and taking the absolute value of the lowest
about 𝑥 where 𝑓(𝑥) is represented by a power series. one minus 𝑥0.
𝑥 is singular at 𝑥 = 6, 9
ex: 𝑥 2 − 15𝑥 + 54 𝑦 ′′ + 𝑥𝑦 ′ − 𝑦 = 0
This means 𝑓(𝑥) needs to have derivatives of all orders at 𝑥0. 𝑥 1
6−0 <𝑅 →6<𝑅
𝑦 ′′ + 𝑦′ − 𝑦=0 Min Radius of
ex: 𝑓 𝑥 = 𝑥 − 1 is not analytic at 𝑥0 = 1 because it is not differentiable at 𝑥0 = 1 𝑥 − 6 (𝑥 − 9) 𝑥 − 6 (𝑥 − 9) Convergence is 6
This also means that the power series representation of 𝑓(𝑥) Phase Plane ex.
is the Taylor series of it. An autonomous system is one where the
𝒅𝒙
= 𝒇(𝒙, 𝒚)
𝒅𝒙
𝒅𝒕
= 𝟑𝒚𝟑
𝒅𝒕
𝒅𝒚
independent variable is not on the right side. 𝒅𝒚
Ordinary vs Singular Points 𝒅𝒕
= 𝒈(𝒙, 𝒚)
𝒅𝒚
𝒅𝒕
=𝒚

To solve for the related phase-plane 𝒅𝒚

= 𝒅𝒕 =
𝒈 𝒙, 𝒚 𝒅𝒚
=
𝟏 1
න𝑦2𝑑𝑦 = න 𝑑𝑥
Given the DEQ: 𝑎2 + 𝑎1 (𝑥)𝑦 ′′
+ 𝑎0(𝑥)𝑦 = 0 (𝑥)𝑦 ′ equation, you can do the following:
𝒅𝒙 𝒅𝒙 𝒇 𝒙, 𝒚
𝒅𝒕
𝒅𝒙 𝟑𝒚𝟐
1 3 1
𝑦 = 𝑥+𝐶
1
3
1
𝐶 = 𝑦3 − 𝑥
and its standard form: 𝑦 + 𝑃 𝑥 𝑦 + 𝑄 𝑥 𝑦 = 0,
′′ ′ Solve for y (probably by 3 3 3 3
separation of variables)

𝑎 𝑥 𝑎 𝑥 To solve for the trajectory of the system (the 𝑪: 𝒙 𝒕 = 𝒙 𝑪: 𝒙 𝒕 = 𝒆𝟑𝒕 = 𝒙

where 𝑃 𝑥 = 𝑎1 (𝑥) and Q 𝑥 = 𝑎0 (𝑥) phase portrait) on the 𝑥𝑦-plane of the
2 2 𝒚 𝒕 =𝒚 𝒚 𝒕 = 𝒆𝒕 = 𝒚
system (the phase plane), solve for a curve
A point is an ordinary point if both 𝑃 𝑥 and Q 𝑥 are analytic at 𝑥0 𝑪: 𝒙 = 𝒚𝟑
(𝐶) that is in terms of only 𝑥 and 𝑦. 𝒚=𝟑 𝒙
A point is a singular point if either 𝑃 𝑥 and Q 𝑥 are not analytic at 𝑥0
A critical point or equilibrium 𝑥0 , 𝑦0 is 𝒇 𝒙𝟎, 𝒚𝟎 = 𝟎
and
(𝑎𝑛𝑦 𝑟𝑒𝑎𝑙 𝑛𝑢𝑚𝑏𝑒𝑟, 0)
Since th ere is no 𝑥 in 𝑓
Solving DEQs with Power Series’ and Ordinary Points one that makes the following true: 𝒈 𝒙𝟎, 𝒚𝟎 = 𝟎 or 𝑔, 𝑥 can be anything.
It’s just the
To solve: 𝑎2 𝑥 𝑦 ′′ + 𝑎1 𝑥 𝑦 ′ + 𝑎0 𝑥 𝑦 = 0, do the following: A graph of the trajectory of the system slope/direction
∞ 𝒅𝒚
field of 𝒅𝒙
1. Let 𝑦 = ෍ 𝑎𝑛 𝑥 − 𝑥0 𝑛

𝑛=0
Types of Phase Plane Points
2. Find 𝑦′ and 𝑦′′ and substitute them into the DEQ A critical point that has most of the A critical point that has the solutions all
(by shifting the solutions flow into it is called go in or out of it is called a node.
3. Simplify and make each summation be in terms of 𝑥 𝑛 sum index)
∞ asymptotically stable. A stable node is called a sink and an
That’s the
4. You’ll end up with ෍ (𝑠𝑡𝑢𝑓𝑓 𝑖𝑛𝑣𝑜𝑙𝑣𝑖𝑛𝑔 𝑎𝑛 ) 𝑥 − 𝑥0 𝑛 = 0 recurrence relation! A critical point that has most of the unstable node is called a source.
𝑛=1 solutions flow away it is called unstable.
5. Set the (𝒔𝒕𝒖𝒇𝒇 𝒊𝒏𝒗𝒐𝒍𝒗𝒊𝒏𝒈 𝒂𝒏) equal to 0 and solve for 𝒂𝒏+𝟏 A critical point that has solutions that lie
on the line of the function that go to the
6. Solve for 𝑎1 , 𝑎2 ... by plugging 𝑛 = 0, 1, 2... into the recurrence relation.
point is called a saddle point.
They will be in terms of 𝑎0 Keep what the original sum in dexes are when you p lug in 𝑛, like if o ne
summation starts at 𝑛 = 2, igno re that term when y ou plu g in 𝑛 = 0 or 1 A critical point that has the solutions
6a. If the problem asks for the 1st 4 terms, multiply your terms by 𝑥𝑛 revolve around it in a circle is called a
center point.
6b. If the problem asks for your function, find what power series
A critical point that has the solutions go
has the same pattern of coefficients. around it in a spiral is called a spiral.
 And with this we are all done!
I just organized the notes a bit better
(more condensed so less half white pages)

It has been a terrible journey and I hope

you enjoyed it (i did not)

send me a text message if you actually

used this (im talking to you jason’s friend)