Higher-Order DEs & Special Methods: Solving, Classifying & Applications

TAKEAWAYS:

• Classification is key: A second-order DE like $y^{\prime \prime }+p(x)y^{\prime }+q(x)y=g(x)$ is linear if $y$ and its derivatives appear to the first power, homogeneous if $g(x)=0$, and constant-coefficient if $p(x),q(x)$ are constants.
• Characteristic equations solve constant-coefficient homogeneous DEs: roots $r_1,r_2$ of $r^2+pr+q=0$ give solutions $y=c_1{e}^{r_{1}x}+c_2{e}^{r_{2}x}$ (real roots) or $y=e^{\alpha x}(c_1\cos \beta x+c_2\sin \beta x)$ (complex roots $\alpha \pm i\beta$).
• Nonhomogeneous terms $g(x)$ require method of undetermined coefficients (for polynomials/exponentials) or variation of parameters (for general $g(x)$). The particular solution $y_p$ depends on the form of $g(x)$.
• Special methods like Bernoulli’s equation ($y^{\prime }+P(x)y=Q(x)y^n$) and exact equations ($M(x,y)dx+N(x,y)dy=0$, where $\frac{\partial M}{\partial y}=\frac{\partial N}{\partial x}$) convert nonlinear DEs into solvable forms.
• Real-world applications include spring-mass systems (harmonic motion), RC circuits (charge/discharge), and population models (logistic growth). Always identify the DE type from the physical law.
• Exam focus: Solve initial value problems (IVPs), classify DEs, and apply variation of parameters or undetermined coefficients to nonhomogeneous equations. Memorize standard forms for $g(x)$ (polynomials, exponentials, sines/cosines).

1. Classification of Higher-Order Differential Equations

Higher-order DEs extend first-order concepts. The general form of a second-order linear DE is: $$y^{\prime \prime }+p(x)y^{\prime }+q(x)y=g(x)$$ where:

• Order: Highest derivative (here, 2).
• Linearity: $y$ and its derivatives appear linearly (no $y^2$, $\sin (y^{\prime })$, etc.).
• Homogeneity: $g(x)=0$ (homogeneous) or $g(x)\ne 0$ (nonhomogeneous).

Visual: Classification Tree

graph TD
    A["Second-Order DE"] --> B{"Linear?"}
    B -->|"Yes"| C{"Constant Coefficients?"}
    C -->|"Yes"| D["Solve via Characteristic Eq."]
    C -->|"No"| E["Solve via Reduction of Order"]
    B -->|"No"| F["Special Methods: Bernoulli, Exact, etc."]
    A -->|"Nonlinear"| G["Transform to Linear or Use Numerical Methods"]

Worked Example 1: Classify and Solve

Problem: Classify and solve $y^{\prime \prime }-3y^{\prime }+2y=0$. Solution:

1. Classification: Linear, homogeneous, constant-coefficient.
2. Characteristic equation: $r^2-3r+2=0$.
3. Roots: $r=1,2$ (real, distinct).
4. General solution: $$y(x)=c_1{e}^x+c_2{e}^{2x}$$

2. Solving Homogeneous Equations with Constant Coefficients

For $y^{\prime \prime }+p{y}^{\prime }+qy=0$, the characteristic equation is: $$r^2+pr+q=0$$ Cases:

1. Distinct real roots $r_1,r_2$: $$y(x)=c_1{e}^{r_{1}x}+c_2{e}^{r_{2}x}$$
2. Repeated real root $r$: $$y(x)=(c_1+c_{2}x)e^{rx}$$
3. Complex roots $\alpha \pm i\beta$: $$y(x)=e^{\alpha x}(c_1\cos \beta x+c_2\sin \beta x)$$

Worked Example 2: Complex Roots

Problem: Solve $y^{\prime \prime }+4y^{\prime }+5y=0$. Solution:

1. Characteristic equation: $r^2+4r+5=0$.
2. Roots: $r=-2\pm i$.
3. General solution: $$y(x)=e^{-2x}(c_1\cos x+c_2\sin x)$$

Real-World Tie: RC Circuit Analysis

Example: In an RC circuit, the charge $Q(t)$ on a capacitor satisfies: $$L\frac{d^{2}Q}{d{t}^2}+R\frac{dQ}{dt}+\frac{1}{C}Q=0$$ This is a second-order DE with constant coefficients. The solution describes damped oscillations (like a ringing circuit after a voltage pulse). Visual:

graph LR
    A["Voltage Source"] --> B["Resistor R"]
    B --> C["Capacitor C"]
    C -->|"Feedback"| A
    label B "Damping"
    label C "Oscillation"

3. Nonhomogeneous Equations: Undetermined Coefficients

For $y^{\prime \prime }+p{y}^{\prime }+qy=g(x)$, the particular solution $y_p$ depends on $g(x)$:

Rule: If $g(x)$ matches a term in $y_h$, multiply by $x$ (e.g., if $g(x)=e^{2x}$ and $e^{2x}$ is in $y_h$, guess $Ax{e}^{2x}$).

Worked Example 3: Undetermined Coefficients

Problem: Solve $y^{\prime \prime }-y^{\prime }-2y=3e^{2x}$. Solution:

1. Homogeneous solution: $y_h=c_1{e}^{2x}+c_2{e}^{-x}$.
2. Guess for $y_p$: Since $3e^{2x}$ matches $c_1{e}^{2x}$ in $y_h$, multiply by $x$: $y_p=Ax{e}^{2x}$.
3. Compute derivatives: $y_p^{\prime }=A(e^{2x}+2x{e}^{2x})$, $y_p^{\prime \prime }=A(4x{e}^{2x}+2e^{2x})$.
4. Substitute into DE: $4Ax{e}^{2x}+2A{e}^{2x}-A(e^{2x}+2x{e}^{2x})-2Ax{e}^{2x}=3e^{2x}$. Simplify: $(2A-A)e^{2x}+(4A-2A)x{e}^{2x}=3e^{2x}$. $A{e}^{2x}=3e^{2x}$ ⇒ $A=3$.
5. General solution: $$y(x)=c_1{e}^{2x}+c_2{e}^{-x}+3x{e}^{2x}$$

Real-World Tie: Ncell’s Signal Attenuation

Ncell’s 4G signal strength in a building follows a nonhomogeneous DE due to:

1. Homogeneous part: Natural decay of signal (like $y^{\prime \prime }+ky=0$).
2. Nonhomogeneous part: External transmitters boosting signal ($g(x)=\text{transmitter power}$). The solution $y_p$ accounts for signal reinforcement in crowded areas.

4. Variation of Parameters

For nonhomogeneous DEs where $g(x)$ is not a simple exponential/polynomial, use variation of parameters: Given $y_h=c_1{y}_1+c_2{y}_2$, the particular solution is: $$y_p=-y_1\int \frac{y_{2}g(x)}{W}dx+y_2\int \frac{y_{1}g(x)}{W}dx$$ where $W=y_1{y}_2^{\prime }-y_2{y}_1^{\prime }$ is the Wronskian.

Worked Example 4: Variation of Parameters

Problem: Solve $y^{\prime \prime }+y=\tan x$ (using $y_1=\sin x$, $y_2=\cos x$). Solution:

1. Compute Wronskian: $W=\sin x(-\sin x)-\cos x(\cos x)=-({\sin }^{2}x+{\cos }^{2}x)=-1$.
2. Particular solution: $$y_p=-\sin x\int \frac{\cos x\tan x}{-1}dx+\cos x\int \frac{\sin x\tan x}{-1}dx$$ Simplify $\tan x=\frac{\sin x}{\cos x}$: $$y_p=\sin x\int \sin xdx-\cos x\int \cos xdx=\sin x(-\cos x)-\cos x(\sin x)=-\sin x\cos x-\sin x\cos x=-2\sin x\cos x$$ (Note: This is simplified; full integration includes constants.)
3. General solution: $$y(x)=c_1\sin x+c_2\cos x-\sin x\cos x$$

Real-World Tie: Pathao’s Ride Demand Forecasting

Pathao’s ride demand $D(t)$ in Kathmandu follows a nonhomogeneous DE: $$\frac{d^{2}D}{d{t}^2}+\alpha \frac{dD}{dt}+\beta D=f(t)$$ where:

• $f(t)$ = external factors (e.g., festivals, traffic jams).
• Variation of parameters helps predict demand spikes when $f(t)$ is complex (e.g., $\tan (t)$-like seasonality).

5. Special Methods

A. Bernoulli’s Equation

Form: $y^{\prime }+P(x)y=Q(x)y^n$. Solution: Substitute $v=y^{1-n}$, transforming it into a linear DE.

Worked Example 5: Bernoulli’s Equation Problem: Solve $y^{\prime }+\frac{y}{x}=x{y}^3$. Solution:

1. Let $v=y^{-2}$, then $v^{\prime }=-2y^{-3}{y}^{\prime }$.
2. Rewrite DE: $-2y^{-3}{y}^{\prime }+\frac{y^{-2}}{x}=x$. Multiply by $-2y^3$: $v^{\prime }-\frac{2}{x}v=-2x$.
3. Solve the linear DE: Integrating factor $\mu (x)=e^{\int -\frac{2}{x}dx}=e^{-2\ln x}=x^{-2}$. Multiply through: $x^{-2}{v}^{\prime }-2x^{-3}v=-2x^{-1}$. Left side is $\frac{d}{dx}(x^{-2}v)$, so: $x^{-2}v=\int -2x^{-1}dx=-2\ln x+C$. Thus: $v=-2x^2\ln x+C{x}^2$. Substitute back $v=y^{-2}$: $y^{-2}=-2x^2\ln x+C{x}^2$.

B. Exact Equations

An equation $M(x,y)dx+N(x,y)dy=0$ is exact if: $$\frac{\partial M}{\partial y}=\frac{\partial N}{\partial x}$$ Solution: Find $\psi (x,y)$ such that $\frac{\partial \psi }{\partial x}=M$ and $\frac{\partial \psi }{\partial y}=N$.

Worked Example 6: Exact Equation Problem: Solve $(2xy+3)dx+(x^2+4y)dy=0$. Solution:

1. Check exactness: $\frac{\partial M}{\partial y}=2x$, $\frac{\partial N}{\partial x}=2x$. Exact!
2. Integrate $M$ w.r.t. $x$: $\psi (x,y)=\int (2xy+3)dx=x^{2}y+3x+h(y)$.
3. Differentiate $\psi$ w.r.t. $y$ and set equal to $N$: $x^2+h^{\prime }(y)=x^2+4y$ ⇒ $h^{\prime }(y)=4y$ ⇒ $h(y)=2y^2$.
4. General solution: $\psi (x,y)=C$ ⇒ $x^{2}y+3x+2y^2=C$.

Real-World Tie: eSewa’s Transaction Fees

eSewa’s transaction fee model can be represented by an exact DE: $$M(x,y)dx+N(x,y)dy=0$$ where:

• $x$ = transaction amount,
• $y$ = fee rate. The exactness condition ensures consistent fee calculation across different transaction sizes.

6. Reduction of Order

For second-order DEs missing $y$ (i.e., $y^{\prime \prime }+p(x)y^{\prime }=0$):

1. Let $v=y^{\prime }$, reducing it to a first-order DE.
2. Solve for $v$, then integrate to find $y$.

Worked Example 7: Reduction of Order Problem: Solve $y^{\prime \prime }-\frac{1}{x}{y}^{\prime }=0$. Solution:

1. Let $v=y^{\prime }$, then $v^{\prime }-\frac{1}{x}v=0$.
2. Solve the first-order DE: $\frac{dv}{dx}=\frac{v}{x}$ ⇒ $\frac{dv}{v}=\frac{dx}{x}$ ⇒ $\ln |v|=\ln |x|+C$ ⇒ $v=Cx$.
3. Integrate to find $y$: $y^{\prime }=Cx$ ⇒ $y=\frac{C}{2}{x}^2+D$.

Real-World Tie: NTC’s Power Distribution

NTC’s power line resistance $R(x)$ along a cable satisfies: $$\frac{d^{2}R}{d{x}^2}-\frac{1}{x}\frac{dR}{dx}=0$$ Reduction of order helps model voltage drop in long-distance transmission lines.

7. Initial Value Problems (IVPs)

Solve the DE with initial conditions (e.g., $y(a)=b$, $y^{\prime }(a)=c$) to find constants $c_1,c_2$.

Worked Example 8: IVP Problem: Solve $y^{\prime \prime }-3y^{\prime }+2y=0$ with $y(0)=1$, $y^{\prime }(0)=0$. Solution:

1. General solution: $y(x)=c_1{e}^x+c_2{e}^{2x}$.
2. Apply $y(0)=1$: $c_1+c_2=1$.
3. Compute $y^{\prime }=c_1{e}^x+2c_2{e}^{2x}$, apply $y^{\prime }(0)=0$: $c_1+2c_2=0$.
4. Solve system: $c_1=-2$, $c_2=3$.
5. Particular solution: $y(x)=-2e^x+3e^{2x}$.

Real-World Tie: Daraz’s Order Fulfillment Queue

Daraz’s order processing queue $Q(t)$ follows: $$\frac{d^{2}Q}{d{t}^2}+5\frac{dQ}{dt}+6Q=120e^{-t}$$ with $Q(0)=0$, $Q^{\prime }(0)=0$.

• Homogeneous solution: Models natural decay of unprocessed orders.
• Particular solution: Accounts for new orders arriving at rate $120e^{-t}$.
• IVP ensures the queue starts empty and grows smoothly.

In the Real World

1. Ncell’s Signal Optimization
   • DE Used: $y^{\prime \prime }+\alpha y^{\prime }+\beta y=g(x)$ (nonhomogeneous).
   • How: The homogeneous part models signal decay in buildings, while $g(x)$ represents transmitter boosts. Solving this DE helps place cell towers to maximize coverage.
   • Visual:

2. Khalti’s Transaction Security

   • DE Used: Bernoulli’s equation $y^{\prime }+P(x)y=Q(x)y^n$.
   • How: Models fraud detection probability $P(t)$ over time, where:
     • $y$ = remaining vulnerable transactions,
     • $Q(x)$ = security patch effectiveness.
   • Example: If $P(t)$ grows nonlinearly, Khalti uses $v=y^{-1}$ to linearize and predict fraud spikes.

3. NEPSE Stock Price Modeling

   • DE Used: Exact equations $M(x,y)dx+N(x,y)dy=0$.
   • How: Represents supply-demand balance for a stock:
     • $M(x,y)=\text{buying pressure}$,
     • $N(x,y)=\text{selling pressure}$.
   • Exactness ensures consistent price predictions when $\frac{\partial M}{\partial y}=\frac{\partial N}{\partial x}$.

Exam Tip

1. Always classify first: Write whether the DE is linear/nonlinear, homogeneous/nonhomogeneous, constant-coefficient/variable-coefficient. Partial marks are often given for correct classification.
2. For constant-coefficient DEs:
   • Write the characteristic equation clearly.
   • Handle repeated roots and complex roots correctly (use $e^{\alpha x}(c_1\cos \beta x+c_2\sin \beta x)$).
3. Nonhomogeneous methods:
   • Undetermined coefficients: Guess $y_p$ based on $g(x)$. Multiply by $x$ if $g(x)$ matches $y_h$.
   • Variation of parameters: Memorize the formula for $y_p$ and compute the Wronskian $W$ carefully.
4. Special methods:
   • Bernoulli: Recognize the form $y^{\prime }+P(x)y=Q(x)y^n$ and substitute $v=y^{1-n}$.
   • Exact equations: Verify $\frac{\partial M}{\partial y}=\frac{\partial N}{\partial x}$ before proceeding.
5. IVPs: Always apply initial conditions to find constants. If you forget, you lose marks even if the general solution is correct.
6. Real-world connections:
   • Physics: Spring-mass systems ($y^{\prime \prime }+{\omega }^{2}y=0$).
   • Biology: Population models ($y^{\prime }=ky$ or logistic growth).
   • Engineering: Circuit analysis, heat transfer.
   • Economics: Supply-demand equilibrium (exact equations).
7. Common pitfalls:
   • Forgetting to multiply by $x$ when $g(x)$ matches $y_h$.
   • Incorrect Wronskian calculation in variation of parameters.
   • Misapplying initial conditions (e.g., using $y(0)$ for $y^{\prime }$).

Final Checklist for Exam Questions:

1. Classify the DE (linear/nonlinear, homogeneous/nonhomogeneous).
2. Solve the homogeneous part first.
3. Find $y_p$ using the appropriate method (undetermined coefficients, variation of parameters, or special methods).
4. Combine $y_h$ and $y_p$ for the general solution.
5. Apply initial conditions (if given) to find constants.
6. Verify your solution by substituting back into the original DE.