Overview
Definition
Variation of parameters: technique to find particular solution (yp) of linear nonhomogeneous second order ODEs of form y'' + p(x)y' + q(x)y = g(x). Uses known solutions y1, y2 of homogeneous equation y'' + p(x)y' + q(x)y = 0.
Purpose
Purpose: obtain yp when method of undetermined coefficients fails or is inconvenient. Flexibility: works for arbitrary g(x) with continuous coefficients.
Historical Context
Developed by Lagrange in 18th century. Classical alternative to undetermined coefficients. Fundamental in advanced ODE solution theory.
Background Concepts
Second Order Linear ODEs
Form: y'' + p(x)y' + q(x)y = g(x). Solutions split into complementary function (yc) + particular solution (yp). Homogeneous equation: g(x) = 0.
Complementary Function
General solution to homogeneous ODE: yc = C1y1 + C2y2, where y1, y2 are linearly independent solutions.
Linear Independence and Wronskian
Two functions y1, y2 linearly independent if Wronskian W(y1,y2)(x) ≠ 0 on interval. Wronskian defined as W = y1 y2' - y2 y1'.
Derivation of the Method
Starting Point
Assume particular solution in form yp = u1(x)y1 + u2(x)y2, where u1, u2 are unknown functions.
Introducing Constraints
To determine u1, u2 uniquely, impose condition u1'y1 + u2'y2 = 0. Simplifies second derivative computation.
Forming System of Equations
Substitute yp into original ODE. Use constraint to get system: u1' y1 + u2' y2 = 0, u1' y1' + u2' y2' = g(x).
Solving for u1', u2'
System linear in u1', u2'. Using Cramer's rule: u1' = -y2 g / W, u2' = y1 g / W, where W is Wronskian.
Integrating to Find u1, u2
Integrate u1', u2' to get u1, u2. Then form yp = u1 y1 + u2 y2.
Step-by-Step Procedure
Step 1: Solve Homogeneous Equation
Find y1, y2 solving y'' + p(x)y' + q(x)y = 0.
Step 2: Compute Wronskian
Calculate W = y1 y2' - y2 y1'. Verify W ≠ 0 on interval.
Step 3: Formulas for u1', u2'
u1' = -y2(x) g(x) / W(x)u2' = y1(x) g(x) / W(x)Step 4: Integrate u1', u2'
Integrate u1', u2' w.r.t x. Add constants if needed (usually absorbed in complementary function).
Step 5: Construct Particular Solution
yp = u1(x) y1(x) + u2(x) y2(x).
Wronskian and Its Role
Definition
Wronskian W(y1,y2)(x) = y1 y2' - y2 y1'. Measures linear independence.
Properties
If W(x0) ≠ 0 at some x0, y1, y2 are linearly independent on interval. W satisfies Abel’s identity.
Abel’s Identity
W(x) = W(x0) exp(-∫p(t) dt), where p(x) is coefficient from y'' + p(x) y' + q(x) y = 0.
Importance in Variation of Parameters
Denominator in formulas for u1', u2'. Nonzero W guarantees unique solution for u1', u2'.
Worked Examples
Example 1: Constant Coefficients
Equation: y'' - 3y' + 2y = e^{2x}. Homogeneous solutions: y1 = e^x, y2 = e^{2x}.
Step 1: Wronskian
W = e^x * 2e^{2x} - e^{2x} * e^x = e^{3x} ≠ 0.
Step 2: Find u1', u2'
u1' = -y2 g / W = -e^{2x} * e^{2x} / e^{3x} = -e^{x}u2' = y1 g / W = e^x * e^{2x} / e^{3x} = 1Step 3: Integrate
u1 = -∫ e^{x} dx = -e^{x} + C1, u2 = ∫ 1 dx = x + C2.
Step 4: Particular solution
yp = u1 y1 + u2 y2 = (-e^{x}) e^{x} + x e^{2x} = -e^{2x} + x e^{2x} = x e^{2x} - e^{2x}.
Example 2: Variable Coefficients
Equation: x^2 y'' - x y' + y = ln x, x > 0. Homogeneous solutions: y1 = x, y2 = x ln x.
Wronskian
W = x * (1 + ln x) - (ln x) * 1 = x.
u1', u2'
u1' = -y2 g / W = -x ln x * ln x / x = - (ln x)^2u2' = y1 g / W = x * ln x / x = ln xIntegrate and form yp
u1 = -∫ (ln x)^2 dx, u2 = ∫ ln x dx. Use integration by parts to evaluate. Then yp = u1 y1 + u2 y2.
Comparison with Other Methods
Undetermined Coefficients
Requires g(x) of special forms (polynomials, exponentials, sines, cosines). Less general than variation of parameters.
Method of Annihilators
Transforms nonhomogeneous ODE to homogeneous by applying differential operator. Efficient for specific g(x), less flexible.
Advantages of Variation of Parameters
Universality: applies to arbitrary g(x) with continuous coefficients. Directly uses homogeneous solutions.
Disadvantages
Integration complexity: integrals for u1, u2 may be difficult or non-elementary. More algebraic steps.
Applications
Physics
Forced oscillations: find particular solutions of driven harmonic oscillator equations.
Engineering
Circuit analysis: solving second order linear ODEs with source terms.
Mathematics
Boundary value problems, Green’s functions construction.
Economics and Biology
Modeling systems with external forcing or inputs.
Limitations and Considerations
Conditions on Coefficients
p(x), q(x), g(x) must be continuous on interval for existence of solutions.
Wronskian Nonzero
Wronskian must not vanish. If zero, y1, y2 not independent, method fails.
Complexity of Integrals
Integrals for u1, u2 may not have elementary forms. Numerical methods may be required.
Extensions to Higher Order ODEs
Generalization
Variation of parameters applicable to nth order linear ODEs. Requires n linearly independent homogeneous solutions.
System of Equations
Form systems for derivatives of parameter functions based on Wronskian matrix inversion.
Computational Complexity
Increases with order: more integrals and algebra involved.
Computational Aspects
Symbolic Computation
CAS tools (Mathematica, Maple, SymPy) automate integrals and algebraic manipulations.
Numerical Integration
If integrals are intractable, numerical quadrature methods approximate u1, u2.
Algorithmic Steps
Input: p(x), q(x), g(x), y1, y2Compute Wronskian WCalculate u1', u2' using formulasIntegrate to find u1, u2Form particular solution yp = u1 y1 + u2 y2Output: General solution y = yc + ypSummary
Variation of parameters: universal method to find particular solutions of nonhomogeneous second order linear ODEs. Requires homogeneous solutions y1, y2; uses Wronskian to solve for parameter functions u1, u2. Advantages: applicability to broad g(x). Limitations: integral complexity, Wronskian nonvanishing requirement. Extensible to higher order ODEs. Important tool in theory and applications of differential equations.
| Key Formula |
|---|
| yp = -y1(x) ∫ (y2(x) g(x) / W(x)) dx + y2(x) ∫ (y1(x) g(x) / W(x)) dx |
References
- Boyce, W.E., DiPrima, R.C., Elementary Differential Equations and Boundary Value Problems, 10th ed., Wiley, 2012, pp. 200-215.
- Bender, C.M., Orszag, S.A., Advanced Mathematical Methods for Scientists and Engineers, McGraw-Hill, 1978, pp. 45-53.
- Ince, E.L., Ordinary Differential Equations, Dover Publications, 1956, pp. 300-320.
- Arnold, V.I., Ordinary Differential Equations, MIT Press, 1973, pp. 75-85.
- Churchill, R.V., Brown, J.W., Fourier Series and Boundary Value Problems, 7th ed., McGraw-Hill, 2006, pp. 110-125.