1.1, 1.2 Flashcards
Integration by Parts Derivation
deriv ( xy ) = x’y + xy’
Integrate both sides
xy (from a to b) = int(x’y + xy’)
int(xy’) = x*y (from a to be) - int(x’y)
Inputs on the left side, outputs on the right
Order of an Ordinary Differential Equation
largest n s.t. x^(n) appears in the equation
an nth order ODE is linear if it can be put in the form:
a.n(t) * x^n + a.n-1(t) * x^n-1 + … + a.0(t) * x = r(t)
AKA on the left are x, x’, x’’ are all mult. by functions of t. On the right is a function of t.
x’’ + tx^2 = e^(-t^2) is linear or nonlinear?
nonlinear b/c of the x^2
Linear or not?
(1+sin(t))x’’ + x’ * e^t = t-x
Linear, because you can move the x to the other side and it fits the form.
How do you verify a solution to an Ordinary Differential Equation
plug in the derivatives to the equation and see if it is still true
What is an IVP problem (initial value problem)?
A differential equation that also has an initial condition, allowing you to solve for the value of C
Fundamental Theorem of Existence and Uniqueness: x’ = f(x,t) , x(t.0) = x.0. Suppose f continuous at (t.0,x.0) and df/dx continuous at (t.0,x.0). Then:
there exists and interval a<t<b containing t.0 such that on this interval x has a unique solution. However, this does not tell you anything about the size of the interval
A first order ODE is in the form:
x’ + p(t) * x = q(t)
A linear equation is called homogeneous if:
if RHS is 0. AKA, r(t) = 0, q(t) = 0
What is the solution to first order homogeneous equations (i.e. x’ + p(t)x = 0)?
Ce^(-P(t)), where P(t) is any antiderivative of p(t). This can be derived using integrating factor
How do you solve first order inhomogeneous solutions (i.e. x’ + p(t) * x = q(t) )?
Integrating factor method!
1. u(t) = int(p(t))
2. Multiply equation by u(t)
3. Integrate to get solution, where LHS will collapse into a product rule.
What is a particular homogeneous solution vs. a homogeneous solution (i.e. c(t + 1)^2 − (t + 1)^2 )?
Particular homogeneous solution does not include C constant (i.e. second half), and is unique up to adding copies of the homogeneous solution to it. AKA, this is a family of solutions and when c=0 that is the particular solution.
When can you use Method of Undetermined Coefficients?
When equation is a constant coefficient equation ODE (i.e. x’ + px = q(t) ) NOT x’ + p(t)x = q(t)
AND
q(t)’s derivatives span finitely many functions
How do you use Method of Undetermined Coefficients?
- Create an Ansatz (guess) that is a constant times each of the spanned functions (i.e. x(t) = At^2 + Bt + C).
- Plug in the guess into the equation and find a system of equations that matches the RHS.
- Solve the system of equations to get the particular solution
- Add the general solution: Ce^-pt