π Lecture 20 β Undetermined Coefficients (Annihilator Operator Approach)
π Overview: This lecture presents the annihilator operator approach for solving non-homogeneous linear differential equations with constant coefficients. It is a systematic method for finding particular solutions when the input function has specific forms like polynomials, exponentials, sines, cosines, and their sums/products.
ποΈ Topics Covered
The lecture covers the method of undetermined coefficients using the annihilator operator, including: the types of input functions suitable for this method, the 8-step solution procedure (from writing the equation to forming the general solution), and several fully worked examples involving polynomial, exponential, and trigonometric forcing functions. Examples range from simple quadratic inputs to complex combinations like ( x \cos x - \cos x ) and ( 10 e^{-2x} \cos x ). The lecture concludes with an exercise set for independent practice.
π Lecture Summary
Undetermined Coefficients (Annihilator Operator Approach)
The method of undetermined coefficients that uses the annihilator operator approach is limited to non-homogeneous linear differential equations that have constant coefficients AND where the function ( g(x) ) has a specific form.
π Definition β Input function forms: The function ( g(x) ) must be one of these types: a constant ( k ), a polynomial, an exponential ( e^{\alpha x} ), trigonometric functions ( \sin(\beta x), \cos(\beta x) ), or finite sums and products of these functions. Otherwise, the method cannot be applied.
19.1 Solution Method
Consider a non-homogeneous linear differential equation with constant coefficients of order ( n ). If ( L ) denotes the differential operator ( L = a_n D^n + a_{n-1} D^{n-1} + \cdots + a_1 D + a_0 ), then the equation is written as ( L(y) = g(x) ).
The method consists of the following 8 steps:
- Step 1: Write the equation as ( L(y) = g(x) ).
- Step 2: Find the complementary solution ( y_c ) by solving the associated homogeneous equation ( L(y) = 0 ).
- Step 3: Operate on both sides with a differential operator ( L_1 ) that annihilates ( g(x) ).
- Step 4: Find the general solution of the higher-order homogeneous equation ( L_1 L(y) = 0 ).
- Step 5: Delete all terms from the solution in Step 4 that are duplicated in ( y_c ).
- Step 6: Form a linear combination ( y_p ) of the remaining terms β this is the form of the particular solution.
- Step 7: Substitute ( y_p ) into ( L(y) = g(x) ), match coefficients, and solve for the unknown constants.
- Step 8: Form the general solution as ( y = y_c + y_p ).
π‘ Why this matters: The annihilator operator provides a systematic, algebraic way to "destroy" the forcing function, turning a non-homogeneous problem into a homogeneous one of higher order, from which the form of the particular solution can be extracted.
π Definition β Annihilator operator: A differential operator that, when applied to a function, yields zero. For example, ( D^3 ) annihilates ( x^2 ) because the third derivative of ( x^2 ) is zero. ( (D-3) ) annihilates ( e^{3x} ). ( (D^2+1) ) annihilates ( \sin x ) and ( \cos x ).
π Formula: Annihilator for ( x^n ) is ( D^{n+1} ). Annihilator for ( e^{\alpha x} ) is ( D - \alpha ). Annihilator for ( \sin(\beta x) ) or ( \cos(\beta x) ) is ( D^2 + \beta^2 ). For products like ( x^2 e^{2x} ), the annihilator is ( (D-2)^3 ).
Example 1 β Solve ( y'' + 3y' + 2y = 4x^2 )
Step 1: Write as ( (D^2 + 3D + 2)y = 4x^2 ). Step 2: Auxiliary equation: ( m^2 + 3m + 2 = (m+1)(m+2) = 0 \Rightarrow m = -1, -2 ). So ( y_c = c_1 e^{-x} + c_2 e^{-2x} ). Step 3: Since ( D^3(4x^2) = 0 ), apply ( D^3 ): ( D^3(D^2 + 3D + 2)y = 0 ). Step 4: Auxiliary equation: ( m^3(m+1)(m+2) = 0 \Rightarrow m = 0,0,0,-1,-2 ). General solution: ( y = c_1 + c_2 x + c_3 x^2 + c_4 e^{-x} + c_5 e^{-2x} ). Step 5: Remove ( c_4 e^{-x} + c_5 e^{-2x} ) (duplicated in ( y_c )). Remaining: ( c_1 + c_2 x + c_3 x^2 ). Step 6: Form ( y_p = A + Bx + Cx^2 ). Step 7: ( y_p' = B + 2Cx ), ( y_p'' = 2C ). Substitute: ( 2C + 3(B + 2Cx) + 2(A + Bx + Cx^2) = 4x^2 ). Simplify: ( 2Cx^2 + (2B + 6C)x + (2A + 3B + 2C) = 4x^2 + 0x + 0 ). Equate coefficients: ( 2C = 4 ), ( 2B + 6C = 0 ), ( 2A + 3B + 2C = 0 ). Solve: ( C = 2 ), ( B = -6 ), ( A = 7 ). So ( y_p = 7 - 6x + 2x^2 ). Step 8: General solution: ( y = c_1 e^{-x} + c_2 e^{-2x} + 7 - 6x + 2x^2 ).
π Example: For ( y = c_1 e^{-x} + c_2 e^{-2x} + 7 - 6x + 2x^2 ), the particular solution ( y_p = 7 - 6x + 2x^2 ) was found by assuming a quadratic form and solving the coefficient equations.
Example 2 β Solve ( y'' - 3y' = 8e^{3x} + 4\sin x )
Step 1: Write as ( (D^2 - 3D)y = 8e^{3x} + 4\sin x ). Step 2: Auxiliary equation: ( m(m-3) = 0 \Rightarrow m = 0, 3 ). So ( y_c = c_1 + c_2 e^{3x} ). Step 3: ( (D-3) ) annihilates ( 8e^{3x} ), ( (D^2+1) ) annihilates ( 4\sin x ). Apply ( (D-3)(D^2+1) ): ( (D-3)(D^2+1)(D^2-3D)y = 0 ). Step 4: Auxiliary: ( m(m-3)^2(m^2+1) = 0 \Rightarrow m = 0, 3, 3, \pm i ). General solution: ( y = c_1 + c_2 e^{3x} + c_3 x e^{3x} + c_4 \cos x + c_5 \sin x ). Step 5: Remove ( c_1 + c_2 e^{3x} ). Remaining: ( c_3 x e^{3x} + c_4 \cos x + c_5 \sin x ). Step 6: Form ( y_p = A x e^{3x} + B \cos x + C \sin x ). Step 7: Substitute into ( y'' - 3y' ). Obtain: ( 3A e^{3x} + (-B - 3C) \cos x + (3B - C) \sin x = 8e^{3x} + 4\sin x ). Equate: ( 3A = 8 ), ( -B - 3C = 0 ), ( 3B - C = 4 ). Solve: ( A = \frac{8}{3} ), ( B = \frac{6}{5} ), ( C = -\frac{2}{5} ). So ( y_p = \frac{8}{3} x e^{3x} + \frac{6}{5} \cos x - \frac{2}{5} \sin x ). Step 8: General solution: ( y = c_1 + c_2 e^{3x} + \frac{8}{3} x e^{3x} + \frac{6}{5} \cos x - \frac{2}{5} \sin x ).
Example 3 β Solve ( y'' + 8y = 5x + 2e^{-x} )
Step 1: Write as ( (D^2 + 8)y = 5x + 2e^{-x} ). Step 2: Auxiliary: ( m^2 + 8 = 0 \Rightarrow m = \pm 2\sqrt{2} i ). So ( y_c = c_1 \cos(2\sqrt{2}x) + c_2 \sin(2\sqrt{2}x) ). Step 3: ( D^2 ) annihilates ( 5x ), ( (D+1) ) annihilates ( 2e^{-x} ). Apply ( D^2(D+1) ): ( D^2(D+1)(D^2+8)y = 0 ). Step 4: Auxiliary: ( m^2(m+1)(m^2+8) = 0 \Rightarrow m = 0,0,-1,\pm 2\sqrt{2} i ). General solution: ( y = c_1 \cos(2\sqrt{2}x) + c_2 \sin(2\sqrt{2}x) + c_3 + c_4 x + c_5 e^{-x} ). Step 5: Remove ( c_1 \cos(2\sqrt{2}x) + c_2 \sin(2\sqrt{2}x) ). Remaining: ( c_3 + c_4 x + c_5 e^{-x} ). Step 6: Form ( y_p = A + Bx + C e^{-x} ). Step 7: Substitute: ( y_p'' + 8y_p = 8A + 8Bx + 9C e^{-x} = 5x + 2e^{-x} ). Equate: ( 8A = 0 ), ( 8B = 5 ), ( 9C = 2 ). Solve: ( A = 0 ), ( B = \frac{5}{8} ), ( C = \frac{2}{9} ). So ( y_p = \frac{5}{8}x + \frac{2}{9} e^{-x} ). Step 8: General solution: ( y = c_1 \cos(2\sqrt{2}x) + c_2 \sin(2\sqrt{2}x) + \frac{5}{8}x + \frac{2}{9} e^{-x} ).
Example 4 β Solve ( y'' + y = x \cos x - \cos x )
Step 1: Write as ( (D^2 + 1)y = x \cos x - \cos x ). Step 2: Auxiliary: ( m^2 + 1 = 0 \Rightarrow m = \pm i ). So ( y_c = c_1 \cos x + c_2 \sin x ). Step 3: ( (D^2+1)^2 ) annihilates ( x \cos x ) and ( \cos x ). Apply ( (D^2+1)^2 ): ( (D^2+1)^3 y = 0 ). Step 4: Auxiliary: ( (m^2+1)^3 = 0 \Rightarrow m = i, i, i, -i, -i, -i ) (each multiplicity 3). General solution: ( y = c_1 \cos x + c_2 \sin x + c_3 x \cos x + c_4 x \sin x + c_5 x^2 \cos x + c_6 x^2 \sin x ). Step 5: Remove ( c_1 \cos x + c_2 \sin x ). Remaining: ( c_3 x \cos x + c_4 x \sin x + c_5 x^2 \cos x + c_6 x^2 \sin x ). Step 6: Form ( y_p = A x \cos x + B x \sin x + C x^2 \cos x + E x^2 \sin x ). Step 7: Substitute into ( y'' + y ): ( y_p'' + y_p = 4Ex \cos x - 4Cx \sin x + (2B + 2C) \cos x + (-2A + 2E) \sin x = x \cos x - \cos x ). Equate: ( 4E = 1 ), ( -4C = 0 ), ( 2B + 2C = -1 ), ( -2A + 2E = 0 ). Solve: ( A = \frac{1}{4} ), ( B = -\frac{1}{2} ), ( C = 0 ), ( E = \frac{1}{4} ). So ( y_p = \frac{1}{4} x \cos x - \frac{1}{2} x \sin x + \frac{1}{4} x^2 \sin x ). Step 8: General solution: ( y = c_1 \cos x + c_2 \sin x + \frac{1}{4} x \cos x - \frac{1}{2} x \sin x + \frac{1}{4} x^2 \sin x ).
Example 5 β Determine the form of a particular solution for ( y'' - 2y' + y = 10e^{-2x} \cos x )
Step 1: Write as ( (D^2 - 2D + 1)y = 10e^{-2x} \cos x ). Step 2: Auxiliary: ( m^2 - 2m + 1 = (m-1)^2 = 0 \Rightarrow m = 1,1 ). So ( y_c = c_1 e^x + c_2 x e^x ). Step 3: ( (D^2 + 4D + 5) ) annihilates ( e^{-2x} \cos x ). Apply: ( (D^2 + 4D + 5)(D^2 - 2D + 1)y = 0 ). Step 4: Auxiliary: ( (m^2 + 4m + 5)(m-1)^2 = 0 \Rightarrow m = -2 \pm i, 1, 1 ). General solution: ( y = c_1 e^x + c_2 x e^x + c_3 e^{-2x} \cos x + c_4 e^{-2x} \sin x ). Step 5: Remove ( c_1 e^x + c_2 x e^x ). Remaining: ( c_3 e^{-2x} \cos x + c_4 e^{-2x} \sin x ). Step 6: Form ( y_p = A e^{-2x} \cos x + B e^{-2x} \sin x ). (Note: Steps 7 and 8 not needed β only the form was requested.)
Example 6 β Determine the form of a particular solution for ( y''' - 4y'' + 4y' = 5x^2 - 6x + 4x^2 e^{2x} + 3e^{5x} )
Step 1: Write as ( (D^3 - 4D^2 + 4D)y = 5x^2 - 6x + 4x^2 e^{2x} + 3e^{5x} ). Step 2: Auxiliary: ( m^3 - 4m^2 + 4m = m(m-2)^2 = 0 \Rightarrow m = 0, 2, 2 ). So ( y_c = c_1 + c_2 e^{2x} + c_3 x e^{2x} ). Step 3: ( D^3 ) annihilates ( 5x^2 - 6x ), ( (D-2)^3 ) annihilates ( x^2 e^{2x} ), ( (D-5) ) annihilates ( 3e^{5x} ). Apply ( D^3(D-2)^3(D-5) ): ( D^4 (D-2)^5 (D-5) y = 0 ). Step 4: Auxiliary: ( m^4 (m-2)^5 (m-5) = 0 \Rightarrow m = 0,0,0,0, 2,2,2,2,2, 5 ). General solution: ( y = c_1 + c_2 x + c_3 x^2 + c_4 x^3 + c_5 e^{2x} + c_6 x e^{2x} + c_7 x^2 e^{2x} + c_8 x^3 e^{2x} + c_9 x^4 e^{2x} + c_{10} e^{5x} ). Step 5: Remove ( c_1 + c_5 e^{2x} + c_6 x e^{2x} ). Remaining: ( c_2 x + c_3 x^2 + c_4 x^3 + c_7 x^2 e^{2x} + c_8 x^3 e^{2x} + c_9 x^4 e^{2x} + c_{10} e^{5x} ). Step 6: Form ( y_p = A x + B x^2 + C x^3 + E x^2 e^{2x} + F x^3 e^{2x} + G x^4 e^{2x} + H e^{5x} ).
β Key Takeaways
The annihilator operator approach is a powerful, step-by-step method for finding particular solutions of non-homogeneous linear differential equations with constant coefficients, provided the forcing function ( g(x) ) is of a specific algebraic or trigonometric type. The core idea is to find a differential operator that "annihilates" ( g(x) ), turning the non-homogeneous equation into a higher-order homogeneous one. The general solution of that higher-order equation contains both the complementary solution and the form of the particular solution. By removing terms duplicated in ( y_c ), the correct form of ( y_p ) is obtained, and its unknown coefficients are found by substituting back into the original equation and matching terms. Mastery of this method requires knowing the annihilators for common functions and carefully handling cases where the forcing function includes terms that already appear in the homogeneous solution.
π§ Quick Revision Questions
- What types of functions ( g(x) ) are suitable for the annihilator operator approach?
- What is the annihilator operator for the function ( x^3 )?
- In Example 1, why was the term ( c_4 e^{-x} + c_5 e^{-2x} ) removed from the general solution of the higher-order equation?
- In Example 2, which operators annihilate ( 8e^{3x} ) and ( 4\sin x ), respectively?
- In Example 4, why is the annihilator for ( x \cos x ) equal to ( (D^2+1)^2 )?
π Lecture 21 β Variation of Parameters
π Overview: This lecture introduces the method of variation of parameters for finding particular solutions of non-homogeneous linear differential equations. Unlike the method of undetermined coefficients, variation of parameters is more general and can be applied to equations with variable coefficients and non-polynomial forcing functions. The method extends from first-order to second-order equations systematically.
ποΈ Topics Covered
The lecture begins by recalling the structure of non-homogeneous linear differential equations and the general solution format. It then derives the variation of parameters method for first-order equations, showing how it reproduces the familiar integrating factor solution. The core of the lecture develops the method for second-order equations, including the derivation of the two conditions on the unknown functions, the use of Cramer's rule with the Wronskian, and a detailed 8-step summary procedure. Three worked examples demonstrate the complete solution process, including handling repeated roots and trigonometric forcing functions.
π Lecture Summary
Recall
A non-homogeneous linear differential equation with constant coefficients is an equation of the form ( a_n \frac{d^n y}{dx^n} + a_{n-1} \frac{d^{n-1} y}{dx^{n-1}} + \cdots + a_1 \frac{dy}{dx} + a_0 y = g(x) ). The general solution is given by General Solution = Complementary Function + Particular Integral. For a first-order linear differential equation ( \frac{dy}{dx} + P(x)y = f(x) ), the general solution is ( y = e^{-\int P dx} \cdot \int e^{\int P dx} f(x) dx + c_1 e^{-\int P dx} ). Here, the second term ( y_c = c_1 e^{-\int P dx} ) is the solution of the associated homogeneous equation, and the first term ( y_p = e^{-\int P dx} \cdot \int e^{\int P dx} f(x) dx ) is a particular solution.
The Variation of Parameters
20.1 First order equation
The method of variation of parameters provides an alternative derivation of the particular solution for first-order equations. Since ( y_1 = e^{-\int P dx} ) solves the homogeneous equation ( \frac{dy}{dx} + P(x)y = 0 ), the general solution is ( y = c_1 y_1(x) ). The variation of parameters consists of finding a function ( u_1(x) ) such that ( y_p = u_1(x) y_1(x) ) is a particular solution of the non-homogeneous equation.
π Definition β Variation of Parameters (first-order): Replace the constant parameter ( c_1 ) in the complementary function with a variable function ( u_1(x) ) and determine ( u_1 ) by substituting into the non-homogeneous equation.
Substituting ( y_p = u_1 y_1 ) into ( \frac{dy}{dx} + P(x)y = f(x) ) gives ( u_1 \left[ \frac{dy_1}{dx} + P(x)y_1 \right] + y_1 \frac{du_1}{dx} = f(x) ). Since ( y_1 ) solves the homogeneous equation, the bracket term is zero, yielding ( y_1 \frac{du_1}{dx} = f(x) ).
π Formula for ( u_1 ): ( \frac{du_1}{dx} = \frac{f(x)}{y_1(x)} ) β Integrate to find ( u_1(x) = \int \frac{f(x)}{y_1(x)} dx = \int e^{\int P dx} \cdot f(x) dx )
Therefore, ( y_p = e^{-\int P dx} \cdot \int e^{\int P dx} f(x) dx ), which is the same result as the integrating factor method, confirming the validity of the variation of parameters approach.
20.2 Second Order Equation
Consider the second-order linear non-homogeneous differential equation in standard form: ( y'' + P(x)y' + Q(x)y = f(x) ), where ( P(x) ), ( Q(x) ), and ( f(x) ) are continuous on some interval ( I ). The associated homogeneous equation is ( y'' + P(x)y' + Q(x)y = 0 ).
Complementary function: Let ( y_1 ) and ( y_2 ) be two linearly independent solutions of the homogeneous equation. Then ( y_c = c_1 y_1(x) + c_2 y_2(x) ). Both ( y_1 ) and ( y_2 ) satisfy the homogeneous equation identically: ( y_1'' + P y_1' + Q y_1 = 0 ) and ( y_2'' + P y_2' + Q y_2 = 0 ).
π‘ Why this matters: The key insight is that for a second-order equation, we need to replace both constants ( c_1 ) and ( c_2 ) with unknown functions ( u_1(x) ) and ( u_2(x) ), giving us two unknowns requiring two equations.
Particular Integral: Replace the parameters ( c_1 ) and ( c_2 ) with unknown variables ( u_1(x) ) and ( u_2(x) ), so the assumed particular integral is ( y_p = u_1(x) y_1(x) + u_2(x) y_2(x) ). We need two equations involving ( u_1 ) and ( u_2 ). One comes from substituting into the differential equation; the other is imposed to simplify derivatives.
The first derivative is ( y_p' = u_1 y_1' + y_1 u_1' + u_2 y_2' + u_2' y_2 = (u_1 y_1' + u_2 y_2') + (u_1' y_1 + u_2' y_2) ). To avoid second derivatives of ( u_1 ) and ( u_2 ), we impose the condition: ( u_1' y_1 + u_2' y_2 = 0 ). Then ( y_p' = u_1 y_1' + u_2 y_2' ).
π Definition β First condition for variation of parameters: ( u_1' y_1 + u_2' y_2 = 0 ) β this condition is imposed to simplify the derivative calculation by eliminating terms involving ( u_1' ) and ( u_2' ) multiplied by the original functions.
Computing the second derivative: ( y_p'' = u_1 y_1'' + u_1' y_1' + u_2 y_2'' + u_2' y_2' ). Substituting ( y_p ), ( y_p' ), and ( y_p'' ) into the non-homogeneous equation and using the fact that ( y_1 ) and ( y_2 ) satisfy the homogeneous equation yields the second condition: ( u_1' y_1' + u_2' y_2' = f(x) ).
π Definition β Second condition for variation of parameters: ( u_1' y_1' + u_2' y_2' = f(x) ) β this condition comes directly from substituting the assumed solution into the original non-homogeneous differential equation.
Thus, ( u_1 ) and ( u_2 ) must satisfy: ( u_1' y_1 + u_2' y_2 = 0 ) and ( u_1' y_1' + u_2' y_2' = f(x) ).
Using Cramer's rule, the solution is: ( u_1' = \frac{W_1}{W} ), ( u_2' = \frac{W_2}{W} ), where ( W ), ( W_1 ), and ( W_2 ) are the determinants:
( W = \begin{vmatrix} y_1 & y_2 \ y_1' & y_2' \end{vmatrix} ), ( W_1 = \begin{vmatrix} 0 & y_2 \ f(x) & y_2' \end{vmatrix} ), ( W_2 = \begin{vmatrix} y_1 & 0 \ y_1' & f(x) \end{vmatrix} )
The determinant ( W ) is the Wronskian of ( y_1 ) and ( y_2 ). Since ( y_1 ) and ( y_2 ) are linearly independent, ( W(y_1(x), y_2(x)) \neq 0 ) for all ( x \in I ). Integrating the expressions for ( u_1' ) and ( u_2' ) gives ( u_1 ) and ( u_2 ), and hence the particular solution.
20.3 Summary of the Method
To solve ( a_2 y'' + a_1 y' + a_0 y = g(x) ) using variation of parameters, follow these steps:
Step 1: Find the complementary function by solving the associated homogeneous equation ( a_2 y'' + a_1 y' + a_0 y = 0 ).
Step 2: If ( y_c = c_1 y_1 + c_2 y_2 ), compute the Wronskian ( W = \begin{vmatrix} y_1 & y_2 \ y_1' & y_2' \end{vmatrix} ).
Step 3: Divide by ( a_2 ) to get standard form ( y'' + P(x)y' + Q(x)y = f(x) ) and identify ( f(x) ).
Step 4: Construct determinants ( W_1 = \begin{vmatrix} 0 & y_2 \ f(x) & y_2' \end{vmatrix} ) and ( W_2 = \begin{vmatrix} y_1 & 0 \ y_1' & f(x) \end{vmatrix} ).
Step 5: Compute ( u_1' = \frac{W_1}{W} ) and ( u_2' = \frac{W_2}{W} ).
Step 6: Integrate: ( u_1 = \int \frac{W_1}{W} dx ) and ( u_2 = \int \frac{W_2}{W} dx ).
Step 7: The particular solution is ( y_p = u_1 y_1 + u_2 y_2 ).
Step 8: The general solution is ( y = y_c + y_p = c_1 y_1 + c_2 y_2 + u_1 y_1 + u_2 y_2 ).
20.3.1 Constants of Integration
We do not need to introduce constants of integration when computing the indefinite integrals in Step 6. If constants ( a_1 ) and ( b_1 ) were added, they would simply merge with ( c_1 ) and ( c_2 ) in the final general solution, providing no new information.
Example 1
Solve: ( y'' - 4y' + 4y = (x+1)e^{2x} )
Step 1: The auxiliary equation is ( m^2 - 4m + 4 = 0 \Rightarrow (m-2)^2 = 0 \Rightarrow m = 2, 2 ). Thus ( y_c = c_1 e^{2x} + c_2 x e^{2x} ).
Step 2: ( y_1 = e^{2x} ), ( y_2 = x e^{2x} ). Compute Wronskian: ( W = \begin{vmatrix} e^{2x} & x e^{2x} \ 2e^{2x} & 2x e^{2x} + e^{2x} \end{vmatrix} = e^{2x}(2xe^{2x}+e^{2x}) - x e^{2x}(2e^{2x}) = 2xe^{4x} + e^{4x} - 2xe^{4x} = e^{4x} \neq 0 )
Step 3: ( f(x) = (x+1)e^{2x} )
Step 4: Construct determinants: ( W_1 = \begin{vmatrix} 0 & x e^{2x} \ (x+1)e^{2x} & 2x e^{2x} + e^{2x} \end{vmatrix} = 0 - (x+1)e^{2x} \cdot x e^{2x} = -(x+1)x e^{4x} ) ( W_2 = \begin{vmatrix} e^{2x} & 0 \ 2e^{2x} & (x+1)e^{2x} \end{vmatrix} = e^{2x} \cdot (x+1)e^{2x} - 0 = (x+1)e^{4x} )
Step 5: ( u_1' = \frac{W_1}{W} = \frac{-(x+1)x e^{4x}}{e^{4x}} = -x^2 - x ) ( u_2' = \frac{W_2}{W} = \frac{(x+1)e^{4x}}{e^{4x}} = x + 1 )
Step 6: Integrate: ( u_1 = \int (-x^2 - x) dx = -\frac{x^3}{3} - \frac{x^2}{2} ) ( u_2 = \int (x + 1) dx = \frac{x^2}{2} + x )
Step 7: Particular solution: ( y_p = \left(-\frac{x^3}{3} - \frac{x^2}{2}\right) e^{2x} + \left(\frac{x^2}{2} + x\right) x e^{2x} ) ( = \left(-\frac{x^3}{3} - \frac{x^2}{2} + \frac{x^3}{2} + x^2\right) e^{2x} = \left(\frac{x^3}{6} + \frac{x^2}{2}\right) e^{2x} )
π Example: ( y_p = \left(\frac{x^3}{6} + \frac{x^2}{2}\right) e^{2x} )
Step 8: General solution: ( y = c_1 e^{2x} + c_2 x e^{2x} + \left(\frac{x^3}{6} + \frac{x^2}{2}\right) e^{2x} )
Example 2
Solve: ( 4y'' + 36y = \csc 3x )
Step 1: Divide by 4: ( y'' + 9y = 0 ). Auxiliary: ( m^2 + 9 = 0 \Rightarrow m = \pm 3i ). Thus ( y_c = c_1 \cos 3x + c_2 \sin 3x ).
Step 2: ( y_1 = \cos 3x ), ( y_2 = \sin 3x ). Compute Wronskian: ( W = \begin{vmatrix} \cos 3x & \sin 3x \ -3\sin 3x & 3\cos 3x \end{vmatrix} = \cos 3x \cdot 3\cos 3x - \sin 3x \cdot (-3\sin 3x) = 3\cos^2 3x + 3\sin^2 3x = 3 )
Step 3: ( f(x) = \frac{1}{4} \csc 3x )
Step 4: Construct determinants: ( W_1 = \begin{vmatrix} 0 & \sin 3x \ \frac{1}{4}\csc 3x & 3\cos 3x \end{vmatrix} = 0 - \frac{1}{4}\csc 3x \cdot \sin 3x = -\frac{1}{4} ) ( W_2 = \begin{vmatrix} \cos 3x & 0 \ -3\sin 3x & \frac{1}{4}\csc 3x \end{vmatrix} = \cos 3x \cdot \frac{1}{4}\csc 3x - 0 = \frac{1}{4} \cdot \frac{\cos 3x}{\sin 3x} = \frac{1}{4} \cot 3x )
Step 5: ( u_1' = \frac{W_1}{W} = \frac{-1/4}{3} = -\frac{1}{12} ) ( u_2' = \frac{W_2}{W} = \frac{(1/4)\cot 3x}{3} = \frac{1}{12} \cot 3x )
Step 6: Integrate: ( u_1 = \int -\frac{1}{12} dx = -\frac{1}{12}x ) ( u_2 = \int \frac{1}{12} \cot 3x dx = \frac{1}{12} \cdot \frac{1}{3} \ln|\sin 3x| = \frac{1}{36} \ln|\sin 3x| )
Step 7: Particular solution: ( y_p = \left(-\frac{1}{12}x\right) \cos 3x + \left(\frac{1}{36} \ln|\sin 3x|\right) \sin 3x )
π Example: ( y_p = -\frac{1}{12}x \cos 3x + \frac{1}{36} (\sin 3x) \ln|\sin 3x| )
Step 8: General solution: ( y = c_1 \cos 3x + c_2 \sin 3x - \frac{1}{12}x \cos 3x + \frac{1}{36} (\sin 3x) \ln|\sin 3x| )
Example 3
Solve: ( y'' - y = \frac{1}{x} )
Step 1: Consider ( y'' - y = 0 ). Auxiliary: ( m^2 - 1 = 0 \Rightarrow m = \pm 1 ). Thus ( y_c = c_1 e^x + c_2 e^{-x} ).
Step 2: ( y_1 = e^x ), ( y_2 = e^{-x} ). Compute Wronskian: ( W = \begin{vmatrix} e^x & e^{-x} \ e^x & -e^{-x} \end{vmatrix} = e^x \cdot (-e^{-x}) - e^{-x} \cdot e^x = -1 - 1 = -2 )
Step 3: ( f(x) = \frac{1}{x} )
Step 4: Construct determinants: ( W_1 = \begin{vmatrix} 0 & e^{-x} \ \frac{1}{x} & -e^{-x} \end{vmatrix} = 0 - \frac{1}{x} \cdot e^{-x} = -\frac{e^{-x}}{x} ) ( W_2 = \begin{vmatrix} e^x & 0 \ e^x & \frac{1}{x} \end{vmatrix} = e^x \cdot \frac{1}{x} - 0 = \frac{e^x}{x} )
Step 5: ( u_1' = \frac{W_1}{W} = \frac{-e^{-x}/x}{-2} = \frac{e^{-x}}{2x} ) ( u_2' = \frac{W_2}{W} = \frac{e^x/x}{-2} = -\frac{e^x}{2x} )
Step 6: Integrate: ( u_1 = \int \frac{e^{-x}}{2x} dx ), ( u_2 = \int -\frac{e^x}{2x} dx )
Step 7: Particular solution: ( y_p = \left(\int \frac{e^{-x}}{2x} dx\right) e^x + \left(\int -\frac{e^x}{2x} dx\right) e^{-x} )
Step 8: General solution: ( y = c_1 e^x + c_2 e^{-x} + y_p )
β Key Takeaways
The variation of parameters method is a powerful and general technique for finding particular solutions to non-homogeneous linear differential equations that works even when the forcing function is not of the special forms required by undetermined coefficients. The procedure for second-order equations involves replacing the constants in the complementary function with unknown functions, imposing the condition ( u_1' y_1 + u_2' y_2 = 0 ) to simplify derivatives, and obtaining a second condition ( u_1' y_1' + u_2' y_2' = f(x) ) from the differential equation itself. The derivatives of the unknown functions are found using Cramer's rule with the Wronskian, and integration (without constants) yields the functions ( u_1 ) and ( u_2 ). The final particular solution is the linear combination ( y_p = u_1 y_1 + u_2 y_2 ), and constants of integration from the indefinite integrals can be omitted as they are absorbed into the complementary function.
π§ Quick Revision Questions
- What two conditions must ( u_1' ) and ( u_2' ) satisfy in the variation of parameters method for a second-order equation?
- Why is the condition ( u_1' y_1 + u_2' y_2 = 0 ) imposed in the derivation?
- What role does the Wronskian play in the variation of parameters method?
- How does the result of variation of parameters for first-order equations compare with the integrating factor method?
- In Example 2, why is the Wronskian a constant (3), and what does this imply about the solutions ( \cos 3x ) and ( \sin 3x )?
π Lecture 22 β Variation of Parameters Method
π Overview: This lecture extends the variation of parameters method for solving non-homogeneous differential equations. It covers the technique for second-order equations with detailed examples, then generalizes the method for higher-order linear differential equations, showing how to find particular solutions when the method of undetermined coefficients is not applicable.
ποΈ Topics Covered
The lecture begins by revisiting the variation of parameters method for second-order equations, demonstrating an example where integrals cannot be expressed in elementary functions. It then generalizes the method to nth-order differential equations, presenting step-by-step procedures for finding complementary functions and particular integrals. Three comprehensive examples are worked out, covering various types of non-homogeneous terms including csc x, tan x, and exponential functions. The section concludes with an exercise set for practice.
π Lecture Summary
Variation of Parameters Method for Second-Order Equations (Review)
The method is demonstrated for solving non-homogeneous equations where the particular integral cannot be found by simple inspection. For the equation y'' - y = 1/x, the process involves finding the complementary function from the homogeneous equation, computing the Wronskian of linearly independent solutions, and then determining particular solutions through integration.
π Definition β Wronskian: For two functions yβ and yβ, the Wronskian W(yβ, yβ) = yβyβ' - yβyβ'. It determines whether solutions are linearly independent if W β 0.
π Formula β For second-order equations: The particular solution is yβ = uβyβ + uβyβ, where uβ' = Wβ/W and uβ' = Wβ/W, with Wβ and Wβ being determinants formed by replacing columns of W.
π Example: For y'' - y = 1/x, the complementary function is y_c = cβeΛ£ + cβeβ»Λ£. The Wronskian W = -2. The particular solution involves integrals β«(eβ»α΅/t)dt and β«(eα΅/t)dt, which cannot be expressed in elementary functions, so they are written as definite integrals from xβ to x.
Variation of Parameters Method for Higher-Order Equations
The method is generalized for nth-order linear differential equations of the form aβyβ½βΏβΎ + aβββyβ½βΏβ»ΒΉβΎ + ... + aβy' + aβy = g(x). The process involves eight systematic steps.
π Definition β Linearly independent solutions: Functions yβ, yβ, ..., yβ are linearly independent if the only solution to cβyβ + cβyβ + ... + cβyβ = 0 is cβ = cβ = ... = cβ = 0.
π Formula β For nth-order equations: The derivatives of unknown functions are uβ' = Wβ/W, where Wβ is obtained by replacing the kth column of the Wronskian with the column [0, 0, ..., 0, f(x)]α΅.
π‘ Why this matters: The first n-1 equations in step 5 are assumptions made to simplify the first n-1 derivatives of yβ. The last equation comes from substituting the particular integral into the original differential equation.
Example 1: Solving y''' + y' = csc x
Step 1: The auxiliary equation mΒ³ + m = 0 gives m = 0, m = Β±i. Therefore, the complementary function is y_c = cβ + cβ cos x + cβ sin x.
Step 2: From y_c, we identify yβ = 1, yβ = cos x, yβ = sin x. The Wronskian is computed and simplified to W = 1 (using row operation Rβ + Rβ).
Step 3: The equation is already in standard form: y''' + 0Β·y'' + y' + 0Β·y = csc x.
Step 4: The determinants Wβ, Wβ, Wβ are found by replacing columns of W with [0, 0, csc x]α΅:
- Wβ = csc x(sinΒ²x + cosΒ²x) = csc x
- Wβ = -cot x
- Wβ = -1
Step 5: The derivatives are uβ' = csc x, uβ' = -cot x, uβ' = -1.
Step 6: Integrating gives uβ = ln|csc x - cot x|, uβ = -ln|sin x|, uβ = -x.
Step 7: The particular solution is yβ = ln|csc x - cot x| - cos xΒ·ln|sin x| - xΒ·sin x.
Step 8: The general solution is y = cβ + cβ cos x + cβ sin x + ln|csc x - cot x| - cos xΒ·ln|sin x| - xΒ·sin x.
Example 2: Solving y''' + y' = tan x
This example follows the same steps with f(x) = tan x. The complementary function remains y_c = cβ + cβ cos x + cβ sin x.
The determinants yield: Wβ = tan x, Wβ = -sin x, Wβ = -sin xΒ·tan x.
Integration produces: uβ = -ln|cos x|, uβ = cos x, uβ = sin x - ln|sec x + tan x|.
The particular solution simplifies to yβ = -ln|cos x| + 1 - sin xΒ·ln|sec x + tan x|.
The general solution can be rewritten as y = dβ + cβ cos x + cβ sin x - ln|cos x| - sin xΒ·ln|sec x + tan x|, where dβ = cβ + 1.
Example 3: Solving y''' - 2y'' - y' + 2y = eΒ³Λ£
Step 1: The auxiliary equation mΒ³ - 2mΒ² - m + 2 = 0 factors to (m - 2)(mΒ² - 1) = 0, giving roots m = 1, 2, -1.
Step 2: The complementary function is y_c = cβeΛ£ + cβeΒ²Λ£ + cβeβ»Λ£. The Wronskian is W = 6eΒ²Λ£ β 0.
Step 4: The determinants are:
- Wβ = -3eβ΄Λ£
- Wβ = 2eΒ³Λ£
- Wβ = eβΆΛ£
Step 5: The derivatives are uβ' = -Β½eΒ²Λ£, uβ' = (1/3)eΛ£, uβ' = (1/6)eβ΄Λ£.
Step 6: Integrating: uβ = -ΒΌeΒ²Λ£, uβ = (1/3)eΛ£, uβ = (1/24)eβ΄Λ£.
Step 7: The particular solution is yβ = -(1/4)eΒ³Λ£ + (1/3)eΒ³Λ£ + (1/24)eΒ³Λ£.
Step 8: The general solution is y = cβeΛ£ + cβeΒ²Λ£ + cβeβ»Λ£ - (1/4)eΒ³Λ£ + (1/3)eΒ³Λ£ + (1/24)eΒ³Λ£.
β Key Takeaways
The variation of parameters method provides a systematic way to find particular solutions for non-homogeneous linear differential equations when the method of undetermined coefficients fails. The core process involves finding the complementary function, computing the Wronskian of linearly independent solutions, and determining particular solutions through integration of ratios of determinants. For higher-order equations, the method requires solving an nΓn system to find derivatives of unknown functions. The particular integral may vary depending on how the resulting integrals are evaluated, and initial value problems must be solved using the general solution, not just the complementary function.
π§ Quick Revision Questions
- For an nth-order differential equation, how many equations must be solved to find the derivatives uβ', uβ', ..., uβ'?
- What is the purpose of the first n-1 equations in the system for finding uβ'?
- In Example 1, why does the Wronskian simplify to 1?
- How does the general solution in Example 2 differ from the complementary function?
- For the equation y''' - 2y'' - y' + 2y = eΒ³Λ£, what are the three roots of the auxiliary equation?