Higher-Order Non-Homogeneous DEs - Theory & Concepts

Higher-Order Non-Homogeneous DEs

$Learning Objectives$

Define a non-homogeneous linear differential equation and its general solution structure.
Apply the Superposition Principle for non-homogeneous equations.
Solve DEs using the Method of Undetermined Coefficients.
Utilize the Annihilator Approach to formalize Undetermined Coefficients.
Solve general non-homogeneous DEs using Variation of Parameters.

A linear differential equation is non-homogeneous if the right-hand side is a non-zero function of $x$ .

$General Solution$

The general solution is the sum of the complementary solution ( $y_c$ , solution to the associated homogeneous equation $L[y]=0$ ) and a particular solution ( $y_p$ , any specific solution to the full non-homogeneous equation $L[y]=f(x)$ ).

y = y_c + y_p

$Superposition Principle$

If the non-homogeneous term $f(x)$ consists of a sum of multiple functions, say $f(x) = f_1(x) + f_2(x) + \dots + f_k(x)$ , then we can find individual particular solutions $y_{p1}, y_{p2}, \dots, y_{pk}$ corresponding to each individual forcing term $f_i(x)$ . The overall particular solution is simply their linear combination:

y_p = y_{p1} + y_{p2} + \dots + y_{pk}

This drastically simplifies solving for complex $f(x)$ terms.

Superposition Builds the Particular Response

For a linear operator, separate additive forcing terms can be solved independently and their particular solutions added.

Split f(x) = f₁(x) + ⋯ + fₖ(x)

Solve L[yₚᵢ] = fᵢ for each term

Add yₚ = yₚ₁ + ⋯ + yₚₖ

Verify L[yₚ] = f and combine with y_c

$Method 1: Undetermined Coefficients$

For a constant-coefficient linear equation, this method applies when $f(x)$ is a polynomial, an exponential, a sine, or a cosine function, or a finite sum/product of supported forms. We make an "educated guess" for the form of $y_p$ based on the form of $f(x)$ and its derivatives.

Form of $f(x)$	Assumed Form of $y_p$
$k$ (Constant)	$A$
$ax^n + \dots + cx + d$ (Polynomial)	$A_n x^n + \dots + A_1 x + A_0$
$ce^{kx}$ (Exponential)	$A e^{kx}$
$c \sin(kx)$ or $c \cos(kx)$	$A \sin(kx) + B \cos(kx)$
$(x^n + \dots)e^{kx}$ (Product)	$(A_n x^n + \dots + A_0)e^{kx}$
$e^{kx} \sin(\omega x)$ or $e^{kx} \cos(\omega x)$	$e^{kx} (A \sin(\omega x) + B \cos(\omega x))$

The Multiplication Rule: Compare the trial space for $y_p$ with the complementary solution $y_c$ . If the trial contains a homogeneous term, multiply the affected trial by the smallest power $x^s$ that removes the overlap. For a forcing exponent whose characteristic root has multiplicity $s$ , this is the corresponding root multiplicity.

Forcing Families and Undetermined-Coefficient Trials

The trial spans the derivatives generated by each supported forcing family; polynomial factors preserve the required degree and mixed trigonometric terms stay paired.

Constant or polynomial Pₘ(x) → general polynomial of degree m

Exponential eᵃˣ → A eᵃˣ

Sine or cosine at β → A sin βx + B cos βx

Pₘ(x)eᵃˣ → degree-m polynomial times eᵃˣ

eᵃˣ sin βx or cos βx → eᵃˣ(A sin βx + B cos βx)

$Annihilator Approach$

An alternative way to formally find $y_p$ is using the Annihilator Approach with differential operators. An operator $L$ annihilates a function $f(x)$ if $L[f(x)] = 0$ .

Common Annihilators (where $D = d/dx$ ):

$f(x) = x^k$ is annihilated by $D^{k+1}$ .
$f(x) = e^{ax}$ is annihilated by $(D - a)$ .
$f(x) = \sin(\beta x)$ or $\cos(\beta x)$ is annihilated by $(D^2 + \beta^2)$ .
$f(x) = e^{\alpha x}\sin(\beta x)$ or $e^{\alpha x}\cos(\beta x)$ is annihilated by $(D^2 - 2\alpha D + (\alpha^2 + \beta^2))$ .

Method: Apply the annihilator to both sides of the non-homogeneous equation $L_{orig}[y] = f(x)$ to get $L_{ann}[L_{orig}[y]] = 0$ , a higher-order homogeneous equation. Solve this new homogeneous equation for its general solution. Discard the terms that belong to the original $y_c$ ; the remaining terms form the structure of $y_p$ . Then, substitute this $y_p$ back into the original non-homogeneous DE to solve for the undetermined coefficients.

$Interactive Simulation$

Interact with the simulation below to explore superposition and forced vibrations for non-homogeneous equations.

Forced Vibrations

y'' + 4y = \cos(\omega t)

Explore the superposition principle: $y(t) = y_h(t) + y_p(t)$ . Observe resonance when forcing frequency $\omega$ matches natural frequency $\omega_0 = 2$ .

Forcing Frequency (

\omega

)1.00 rad/s

Toggle Components

Total Solution

y(t)

Homogeneous

y_h(t)

Particular

y_p(t)

max_y ≈ 1.0

$Method 2: Variation of Parameters$

For a normalized second-order linear DE $y'' + P(x)y' + Q(x)y = f(x)$ with continuous coefficients on an interval, this method applies when a known independent fundamental pair $y_1,y_2$ is available there, so its Wronskian $W$ is nonzero. It accommodates forcing outside the undetermined-coefficients family when the required integrals exist. The particular solution takes the form $y_p = u_1(x)y_1(x) + u_2(x)y_2(x)$ .

Variation of Parameters Reuses the Homogeneous Basis

A particular response is formed from the known homogeneous basis by allowing its constant coefficients to vary with x; the Wronskian controls the coefficient derivatives.

Homogeneous solution

y_c = C₁y₁ + C₂y₂
C₁ and C₂ are constants
y₁ and y₂ are independent on the interval

Particular response

yₚ = u₁(x)y₁ + u₂(x)y₂
u₁′ = −y₂f/W; u₂′ = y₁f/W
Integrate where W ≠ 0

Variation of Parameters Formulas

Calculates the coefficient functions $u_1$ and $u_2$ required for the particular solution.

u_1 = \int \frac{-y_2 f(x)}{W} \, dx \quad \text{and} \quad u_2 = \int \frac{y_1 f(x)}{W} \, dx

Variables

Symbol	Description	Unit
$u_1, u_2$	Unknown function coefficients for the particular solution	varies
$y_1, y_2$	Linearly independent solutions to the complementary homogeneous equation	varies
$f(x)$	The non-homogeneous forcing function	varies
$W$	The Wronskian of $y_1$ and $y_2$	varies

$Wronskian Properties$

The Wronskian, denoted as $W(y_1, y_2, \dots, y_n)$ , determines whether a set of solutions to a homogeneous differential equation is linearly independent:

If $W \neq 0$ at any point in the interval, the functions are linearly independent.
If $W = 0$ everywhere in the interval, and the functions are solutions to a regular linear ODE, they are linearly dependent.
Abel's Identity states that $W(x) = W(x_0)e^{-\int_{x_0}^x P(t)dt}$ . This means the Wronskian is either never zero or always zero on an interval where $P(x)$ is continuous.

Selecting and Verifying a Particular Solution

Particular-Solution Method and Resonance Check

Choose a particular-solution method for a linear non-homogeneous ODE, correct undetermined-coefficient trial overlap, and verify the result.

This chart assumes a linear non-homogeneous ODE on a stated interval. Undetermined coefficients applies to constant-coefficient equations when the forcing is a finite sum of polynomial, exponential, sinusoidal, or supported products of these forms. Construct the complete trial for each forcing term, including polynomial factors and paired sine/cosine terms as needed. If a trial term lies in the complementary solution, multiply the affected trial by x and repeat; for a characteristic root of multiplicity s, the required factor is x^s. If undetermined coefficients does not apply, variation of parameters is available here only for a normalized second-order equation y″ + P(x)y′ + Q(x)y = f(x), with a known independent fundamental pair y₁,y₂ and Wronskian W ≠ 0 on the interval. Otherwise select an order-appropriate alternative. Verify the particular solution in the original equation before combining it with y_c.

Key Takeaways

General Solution Form: Always write the solution as $y = y_c + y_p$ . Find $y_c$ first.
Superposition Principle: If $f(x) = f_1(x) + f_2(x)$ , solve for $y_{p1}$ and $y_{p2}$ separately and sum them.
Undetermined Coefficients: Fast and effective for polynomial, exponential, and sinusoidal forcing functions, including their sums and products. The assumed form of $y_p$ mirrors $f(x)$ .
The Multiplication Rule: Crucial step. If your guessed $y_p$ contains terms from $y_c$ , multiply the guess by $x$ (or $x^2$ ) until no duplication exists.
Variation of Parameters: For the normalized second-order equation in this lesson, it uses a known independent pair and nonzero Wronskian on the interval; the required integrals need to exist.

$References$

Previous TopicHigher-Order Homogeneous DEs - Examples & Applications

Quiz Me

Next TopicHigher-Order Non-Homogeneous DEs - Examples & Applications

Prev Next

Quiz Me