📘 Lecture 14 — Boundary-Value Problems and Linear Dependence
📖 Overview: This lecture introduces boundary-value problems (BVPs) as distinct from initial-value problems (IVPs), discussing their unique existence and uniqueness characteristics. It also establishes the fundamental concepts of linear dependence and independence of functions, along with the Wronskian as a tool for determining independence.
🗂️ Topics Covered
The lecture covers the existence and uniqueness of solutions for initial-value problems, the formulation and possible outcomes of boundary-value problems, the definition and examples of linear dependence and independence of functions, the concept of the Wronskian determinant, and the theorem for using it as a criterion for linear independence.
📝 Lecture Summary
13.2.1 Solution of IVP
A function satisfying a differential equation on an interval I whose graph passes through (x₀, y₀) such that the slope of the curve at that point equals y₀′ is called a solution of the initial-value problem.
13.3 Theorem (Existence and Uniqueness of Solutions)
Let aₙ(x), aₙ₋₁(x), …, a₁(x), a₀(x) and g(x) be continuous on an interval I with aₙ(x) ≠ 0 for all x in I. If x₀ is in I, then a solution y(x) of the initial-value problem exists on I and is unique.
🔑 Definition — Initial-Value Problem (IVP): A differential equation together with specified values of the function and its derivatives at a single point x₀.
📌 Example 1: The function y = 3e²ˣ + e⁻²ˣ – 3x solves the IVP y″ – 4y = 12x, y(0)=4, y′(0)=1. Because the equation is linear with constant coefficients, and a₂(x)=1≠0 for all x, this solution is unique.
📌 Example 2: The IVP 3y‴ + 5y″ – y′ + 7y = 0, y(1)=0, y′(1)=0, y″(1)=0 has only the trivial solution y=0, which is unique.
💡 Why this matters: If the leading coefficient aₙ(x)=0 at some point in I, the solution of an IVP may not be unique or may not even exist.
📌 Example 4: The function y = cx² + x + 3 satisfies the IVP x²y″ – 2xy′ + 2y = 6, y(0)=3, y′(0)=1 for any choice of c. The leading coefficient a₂(x)=x²=0 at x=0, causing non-uniqueness.
13.4 Boundary-value problem (BVP)
For a 2ⁿᵈ-order linear differential equation, the problem:
- Solve: a₂(x) d²y/dx² + a₁(x) dy/dx + a₀(x) y = g(x)
- Subject to: y(a)=y₀, y(b)=y₁ is called a boundary-value problem (BVP). The specified values y(a)=y₀ and y(b)=y₁ are called boundary conditions.
13.4.1 Solution of BVP
A solution of the BVP is a function satisfying the differential equation on an interval I containing a and b, whose graph passes through the two points (a, y₀) and (b, y₁).
📌 Example 5: The function y = 3x² – 6x + 3 satisfies the BVP x²y″ – 2xy′ + 2y = 6, y(1)=0, y(2)=3, as verified by substitution.
13.4.2 Possible Boundary Conditions
For a 2ⁿᵈ-order linear differential equation, the possible pairs of boundary conditions include: y(a) = y₀, y(b) = y₁; y′(a) = y₀′, y(b) = y₁; y(a) = y₀, y′(b) = y₁′; y′(a) = y₀′, y′(b) = y₁′. These are special cases of the general form: α₁y(a) + β₁y′(a) = γ₁, α₂y(b) + β₂y′(b) = γ₂, where α₁, α₂, β₁, β₂ ∈ {0,1}.
💡 Why this matters: A BVP can have several solutions, a unique solution, or no solution at all.
📌 Example 1: The BVP y″ + 16y = 0, y(0)=0, y(π/2)=0 has the one-parameter family of solutions y = c₂ sin 4x, meaning infinite solutions. The condition y(π/2)=0 is satisfied for any c₂ because sin(2π)=0.
📌 Example 2: The BVP y″ + 16y = 0, y(0)=0, y(π/8)=0 has only the trivial solution y = 0, because c₁ = 0 from y(0)=0 and c₂ = 0 from y(π/8)=0.
📌 Example 3: The BVP y″ + 16y = 0, y(0)=0, y(π/2)=1 has no solution. Applying y(π/2)=1 to y = c₂ sin 4x yields 1 = c₂·0, a contradiction.
13.5 Linear Dependence
A set of functions {f₁(x), f₂(x), …, fₙ(x)} is linearly dependent on an interval I if there exist constants c₁, c₂, …, cₙ, not all zero, such that c₁ f₁(x) + c₂ f₂(x) + … + cₙ fₙ(x) = 0 for all x in I.
13.6 Linear Independence
A set of functions {f₁(x), f₂(x), …, fₙ(x)} is linearly independent on an interval I if the equation c₁ f₁(x) + c₂ f₂(x) + … + cₙ fₙ(x) = 0 for all x in I implies c₁ = c₂ = … = cₙ = 0.
13.6.1 Case of two functions
Two functions f₁(x) and f₂(x) are linearly dependent on an interval I if and only if one is a constant multiple of the other. They are linearly independent when neither is a constant multiple of the other. In general, a set of n functions is linearly dependent if at least one can be expressed as a linear combination of the rest.
📌 Example 1: f₁(x) = sin 2x and f₂(x) = sin x cos x are linearly dependent because sin 2x = 2 sin x cos x, so (1/2) sin 2x – sin x cos x = 0 with non-zero coefficients (c₁=1/2, c₂=-1).
📌 Example 2: f₁(x) = cos²x, f₂(x) = sin²x, f₃(x) = sec²x, f₄(x) = tan²x are linearly dependent because sec²x = cos²x + sin²x + tan²x, so the combination with c₁=1, c₂=1, c₃=-1, c₄=1 gives zero.
📌 Example 3: f₁(x) = 1 + x, f₂(x) = x, f₃(x) = x² are linearly independent because the equation c₁(1+x) + c₂x + c₃x² = 0 forces c₁=0, c₂=0, c₃=0 when coefficients of like powers are equated.
13.7 Wronskian
If functions f₁, f₂, …, fₙ possess at least n–1 derivatives, the Wronskian, denoted W(f₁, f₂, …, fₙ), is the determinant:
| f₁ f₂ … fₙ |
| f₁′ f₂′ … fₙ′ |
| ⁝ ⁝ ⋱ ⁝ |
| f₁⁽ⁿ⁻¹⁾ f₂⁽ⁿ⁻¹⁾ … fₙ⁽ⁿ⁻¹⁾ |
🔑 Definition — Wronskian: A determinant used to test for linear independence of a set of functions.
📐 Formula: For two functions f₁ and f₂, W = f₁ f₂′ – f₂ f₁′.
13.8 Theorem (Criterion for Linearly Independent Functions)
If functions f₁, f₂, …, fₙ possess at least n–1 derivatives on an interval I and W(f₁, f₂, …, fₙ) ≠ 0 for at least one point in I, then the functions are linearly independent on I. If the functions are linearly dependent on I, then W = 0 for all x in I (the converse is not necessarily true).
📌 Example 1: f₁(x) = sin²x and f₂(x) = 1 – cos 2x are linearly dependent, and their Wronskian W = 0 for all x. This confirms the vanishing Wronskian for dependent functions.
📌 Example 2: f₁(x) = e^(m₁x) and f₂(x) = e^(m₂x) with m₁ ≠ m₂ are linearly independent (neither is a constant multiple of the other). Their Wronskian W = (m₂ – m₁)e^((m₁+m₂)x) ≠ 0.
📌 Example 3: The functions y₁ = e^(αx) cos βx and y₂ = e^(αx) sin βx are linearly independent for β ≠ 0 on any interval, with W = βe^(2αx) ≠ 0.
📌 Example 4: The functions f₁(x) = eˣ, f₂(x) = xeˣ, f₃(x) = x²eˣ are linearly independent on any interval, with W = 2e^(3x) ≠ 0.
⭐ Key Takeaways
You must understand the fundamental difference between initial-value problems (which have a unique solution when the leading coefficient never vanishes) and boundary-value problems (which may have zero, one, or infinitely many solutions). The concept of linear dependence is that one function can be written as a combination of others, while independence requires that only the trivial combination yields zero. The Wronskian provides a practical computational tool: if it is non-zero at any point, the functions are linearly independent. Remember, a zero Wronskian does not guarantee dependence—it is only a sufficient condition for independence.
🧠 Quick Revision Questions
- State the conditions for a unique solution to an initial-value problem.
- What is a boundary-value problem, and how does it differ from an initial-value problem?
- What are the three possible outcomes when solving a boundary-value problem?
- Define linear dependence and linear independence for a set of functions.
- If the Wronskian of two functions is non-zero for all x, what can you conclude?
📘 Lecture 14 — Solutions of Higher Order Linear Equations
📖 Overview: This lecture introduces the theory for solving higher-order linear differential equations, beginning with homogeneous equations. It covers the superposition principle, linear independence, the Wronskian, fundamental sets of solutions, and the structure of general solutions for both homogeneous and non-homogeneous equations.
🗂️ Topics Covered
This lecture covers the preliminary theory for solving nth-order linear differential equations, starting with the homogeneous case. Key concepts include the superposition principle for solutions, the Wronskian's role in determining linear independence, and the definition of a fundamental set of solutions. It also explains the general solution structure for homogeneous equations and introduces the concepts of particular solutions and complementary functions for non-homogeneous equations.
📝 Lecture Summary
14.1 Preliminary Theory
To solve an nth-order non-homogeneous linear differential equation, we first solve the associated homogeneous differential equation. The focus is on the theory and methods for solving homogeneous linear equations. A function that satisfies the homogeneous equation is called a solution.
14.2 Superposition Principle
If ( y_1, y_2, ..., y_n ) are solutions on an interval ( I ) of the homogeneous linear differential equation, then the linear combination ( y = c_1 y_1(x) + c_2 y_2(x) + ... + c_n y_n(x) ), where ( c_1, c_2, ..., c_n ) are arbitrary constants, is also a solution. A constant multiple of a solution is also a solution. Homogeneous linear equations always have the trivial solution ( y = 0 ). This principle does not hold for non-linear equations.
💡 Why this matters: The Superposition Principle is the foundation for constructing general solutions from a set of individual solutions.
🔑 Definition — Superposition Principle: If ( y_1, y_2, ..., y_n ) are solutions to a homogeneous linear DE, then any linear combination of them is also a solution.
📌 Example 1: The functions ( y = e^x, y = e^{2x}, y = e^{3x} ) all satisfy ( y''' - 6y'' + 11y' - 6y = 0 ). Substituting the linear combination ( y = c_1 e^x + c_2 e^{2x} + c_3 e^{3x} ) into the DE, the result is 0, confirming it is also a solution.
📌 Example 2: The function ( y = x^2 ) is a solution of ( x^2 y'' - 3xy' + 4y = 0 ) on ( (0, \infty) ). Substituting ( y = cx^2 ) (a scalar multiple) also satisfies the equation, confirming it is a solution.
The Wronskian
For two solutions ( y_1, y_2 ) of a second-order homogeneous linear DE on interval ( I ), the Wronskian ( W(y_1, y_2) ) is either identically zero or never zero on ( I ). This is derived by showing the Wronskian satisfies the linear first-order DE ( \frac{dW}{dx} + PW = 0 ), leading to ( W = ce^{-\int P dx} ). If ( c \neq 0 ), ( W \neq 0 ); if ( c = 0 ), ( W = 0 ). This property extends to n solutions of an nth-order homogeneous linear DE.
🔑 Definition — Wronskian (for two functions): ( W(y_1, y_2) = y_1 y_2' - y_1' y_2 )
📐 Formula: ( W(y_1, y_2) = \begin{vmatrix} y_1 & y_2 \ y_1' & y_2' \end{vmatrix} ) → The value of this 2x2 determinant determines if the functions are linearly dependent or independent.
14.3 Linear Independence of Solutions
A set of n solutions ( y_1, y_2, ..., y_n ) of a homogeneous linear nth-order DE is linearly independent on interval ( I ) if and only if ( W(y_1, y_2, ..., y_n) \neq 0 ). Conversely, the solutions are linearly dependent if and only if ( W(y_1, y_2, ..., y_n) = 0 ) for all ( x \in I ).
14.4 Fundamental Set of Solutions
A set ( { y_1, y_2, ..., y_n } ) of n linearly independent solutions of a homogeneous linear nth-order DE on interval ( I ) constitutes a fundamental set of solutions.
14.4.1 Existence of a Fundamental Set
A fundamental set of solutions always exists for a linear nth-order homogeneous differential equation on any interval ( I ).
14.5 General Solution-Homogeneous Equations
If ( { y_1, y_2, ..., y_n } ) is a fundamental set of solutions on interval ( I ) for a homogeneous linear nth-order DE, then the general solution on ( I ) is defined as ( y = c_1 y_1(x) + c_2 y_2(x) + ... + c_n y_n(x) ), where ( c_1, c_2, ..., c_n ) are arbitrary constants.
📌 Example 1: For ( y'' - 9y = 0 ), the functions ( y_1 = e^{3x} ) and ( y_2 = e^{-3x} ) are solutions. Since ( W(e^{3x}, e^{-3x}) = -6 \neq 0 ), they form a fundamental set. The general solution is ( y = c_1 e^{3x} + c_2 e^{-3x} ).
📌 Example 2: The function ( y = 4 \sinh 3x - 5e^{-3x} ) satisfies ( y'' - 9y = 0 ) and is a particular solution. Choosing ( c_1 = 2, c_2 = -7 ) from the general solution gives ( y = 2e^{3x} - 7e^{-3x} ), which is equal to ( 4 \sinh 3x - 5e^{-3x} ). This demonstrates that a particular solution can be obtained from the general solution.
📌 Example 3: For ( y''' - 6y'' + 11y' - 6y = 0 ), the functions ( y_1 = e^x, y_2 = e^{2x}, y_3 = e^{3x} ) are all solutions. The Wronskian ( W(e^x, e^{2x}, e^{3x}) = 2e^{6x} \neq 0 ), so they form a fundamental set. The general solution is ( y = c_1 e^x + c_2 e^{2x} + c_3 e^{3x} ).
14.6 Non-Homogeneous Equations
A function ( y_p ) that satisfies the non-homogeneous differential equation and contains no arbitrary parameters is called a particular solution.
📌 Example 1: ( y_p = 3 ) is a particular solution of ( y'' + 9y = 27 ).
📌 Example 2: ( y_p = x^3 - x ) is a particular solution of ( x^2 y'' + 2xy' - 8y = 4x^3 + 6x ).
14.7 Complementary Function
The general solution of the associated homogeneous equation, ( y_c = c_1 y_1 + c_2 y_2 + ... + c_n y_n ), is called the complementary function for the non-homogeneous equation.
⭐ Key Takeaways
The Superposition Principle is crucial for building general solutions of homogeneous linear DEs from individual solutions. The Wronskian is the definitive test for linear independence of solutions: a non-zero Wronskian indicates a fundamental set, which is necessary to form the general solution. The general solution of a homogeneous linear DE is a linear combination of a fundamental set of solutions. For non-homogeneous equations, the general solution includes the complementary function (solution to the homogeneous part) plus any particular solution.
🧠 Quick Revision Questions
- State the Superposition Principle for homogeneous linear differential equations.
- What is the Wronskian used for in the context of solutions to differential equations?
- What condition must a set of solutions satisfy to be considered a fundamental set?
- What is the relationship between the general solution of a homogeneous equation and the general solution of the corresponding non-homogeneous equation?
- Why does the Superposition Principle not apply to non-linear differential equations?
📘 Lecture 14 — General Solution of Non-Homogeneous Equations
📖 Overview: This lecture explains how to construct the general solution of non-homogeneous linear differential equations by combining the complementary function and a particular solution. It also introduces the superposition principle for non-homogeneous equations, which allows us to break down complicated forcing functions into simpler components. Understanding this structure is essential for solving real-world problems where external forces or inputs are present.
🗂️ Topics Covered
The lecture begins by defining the general solution of non-homogeneous equations as the sum of the complementary function and a particular solution, illustrated with a worked example. It then presents the superposition principle, which states that the sum of particular solutions for individual forcing terms is a particular solution for the sum of those terms. Finally, the lecture provides a set of exercises for verifying fundamental solution sets and general solutions.
📝 Lecture Summary
14.8 General Solution of Non-Homogeneous Equations
Suppose that:
- The particular solution of the non-homogeneous equation aₙ(x) dⁿy/dxⁿ + aₙ₋₁(x) dⁿ⁻¹y/dxⁿ⁻¹ + … + a₁(x) dy/dx + a₀(x) y = g(x) is yₚ.
- The complementary function of the associated homogeneous equation aₙ(x) dⁿy/dxⁿ + aₙ₋₁(x) dⁿ⁻¹y/dxⁿ⁻¹ + … + a₁(x) dy/dx + a₀(x) y = 0 is y_c = c₁y₁ + c₂y₂ + … + cₙyₙ.
Then the general solution of the non-homogeneous equation on the interval I is given by: y = y_c + yₚ or y = c₁y₁(x) + c₂y₂(x) + … + cₙyₙ(x) + yₚ(x) = y_c(x) + yₚ(x)
Hence, General Solution = Complementary solution + any particular solution.
🔑 Definition — Complementary Function: The general solution of the associated homogeneous differential equation (where the right-hand side is zero).
📌 Example: Suppose yₚ = -11/12 - (1/2)x. Then yₚ' = -1/2, yₚ'' = 0. Substituting into d³y/dx³ - 6 d²y/dx² + 11 dy/dx - 6y gives 0 - 0 - 11/2 + 11/2 + 3x = 3x. Hence yₚ = -11/12 - (1/2)x is a particular solution of d³y/dx³ - 6 d²y/dx² + 11 dy/dx - 6y = 3x.
Now consider y_c = c₁eˣ + c₂e²ˣ + c₃e³ˣ. Computing y_c', y_c'', y_c''' and substituting into the homogeneous equation d³y/dx³ - 6 d²y/dx² + 11 dy/dx - 6y = 0 yields 0, confirming y_c is the complementary function. Therefore, the general solution of the non-homogeneous equation is: y = y_c + yₚ = c₁eˣ + c₂e²ˣ + c₃e³ˣ - 11/12 - (1/2)x.
14.9 Superposition Principle for Non-homogeneous Equations
Suppose that yₚ₁, yₚ₂, …, yₚₖ denote the particular solutions of the k differential equations aₙ(x)y⁽ⁿ⁾ + aₙ₋₁(x)y⁽ⁿ⁻¹⁾ + … + a₁(x)y' + a₀(x)y = gᵢ(x), for i = 1, 2, …, k, on an interval I. Then yₚ = yₚ₁(x) + yₚ₂(x) + … + yₚₖ(x) is a particular solution of aₙ(x)y⁽ⁿ⁾ + aₙ₋₁(x)y⁽ⁿ⁻¹⁾ + … + a₁(x)y' + a₀(x)y = g₁(x) + g₂(x) + … + gₖ(x).
🔑 Definition — Superposition Principle for Non-homogeneous Equations: The sum of particular solutions corresponding to individual forcing functions is a particular solution of the equation with the sum of those forcing functions.
📌 Example: Consider y'' - 3y' + 4y = -16x² + 24x - 8 + 2e²ˣ + 2xeˣ - eˣ. Suppose yₚ₁ = -4x², yₚ₂ = e²ˣ, yₚ₃ = xeˣ. Then yₚ₁'' - 3yₚ₁' + 4yₚ₁ = -8 + 24x - 16x², so yₚ₁ is a particular solution of y'' - 3y' + 4y = -16x² + 24x - 8. Similarly, yₚ₂ = e²ˣ is a particular solution of y'' - 3y' + 4y = 2e²ˣ, and yₚ₃ = xeˣ is a particular solution of y'' - 3y' + 4y = 2xeˣ - eˣ. Hence, by the superposition principle, yₚ = yₚ₁ + yₚ₂ + yₚ₃ = -4x² + e²ˣ + xeˣ is a particular solution of the original equation.
14.10 Exercise
The lecture provides exercises for verifying that given functions form a fundamental set of solutions on an indicated interval and for forming the general solution. Examples include:
- y'' - y' - 12y = 0; e⁻³ˣ, e⁴ˣ, (-∞, ∞)
- y'' - 2y' + 5y = 0; eˣ cos 2x, eˣ sin 2x, (-∞, ∞)
- x²y'' + xy' + y = 0; cos(ln x), sin(ln x), (0, ∞)
- 4y'' - 4y' + y = 0; eˣ/², xeˣ/², (-∞, ∞)
- x²y'' - 6xy' + 12y = 0; x³, x⁴, (0, ∞)
- y'' - 4y = 0; cosh 2x, sinh 2x, (-∞, ∞)
Additionally, exercises ask to verify that given two-parameter families are the general solution of non-homogeneous differential equations on indicated intervals.
⭐ Key Takeaways
The central concept of this lecture is that the general solution of any non-homogeneous linear differential equation is the sum of the complementary function (the general solution of the associated homogeneous equation) and any particular solution of the non-homogeneous equation. The superposition principle is a powerful tool that allows us to find a particular solution for a complicated forcing function by breaking it into simpler parts, finding a particular solution for each part, and summing them. For exam preparation, you must be able to identify the complementary function from the homogeneous equation, verify if a given function is a particular solution, apply the superposition principle, and write the general solution. The exercises reinforce the method of verifying fundamental solution sets and general solutions for both homogeneous and non-homogeneous equations.
🧠 Quick Revision Questions
- What is the general solution of a non-homogeneous linear differential equation, and what two components make it up?
- State the superposition principle for non-homogeneous differential equations in your own words.
- If yₚ₁ is a particular solution for g₁(x) and yₚ₂ is a particular solution for g₂(x), what is a particular solution for g₁(x) + g₂(x)?
- Given a non-homogeneous equation, how do you find its complementary function?
- In the example, why is y_c = c₁eˣ + c₂e²ˣ + c₃e³ˣ the complementary function for the equation d³y/dx³ - 6 d²y/dx² + 11 dy/dx - 6y = 3x?
📘 Lecture 17 — Method of Undetermined Coefficients (Superposition Approach)
📖 Overview: This lecture introduces a method for finding particular solutions of non-homogeneous linear differential equations with constant coefficients. The method of undetermined coefficients provides a systematic way to guess the form of a particular solution based on the input function ( g(x) ), and its importance lies in being a straightforward technique for equations where ( g(x) ) has a specific form.
🗂️ Topics Covered
This lecture begins by recalling the structure of non-homogeneous linear differential equations and the need for both a complementary function and a particular integral. It then discusses the limitations and specific forms of the input function ( g(x) ) that make the method of undetermined coefficients applicable. The core of the lecture details the step-by-step solution process, emphasizing how to assume the form of the particular solution ( y_p ) and the critical caution about avoiding duplication with terms in the complementary function ( y_c ).
📝 Lecture Summary
17 Method of Undetermined Coefficients (Superposition Approach)
This section introduces the context for solving non-homogeneous linear differential equations. It recalls that the general solution is the sum of the complementary function ( y_c ) (the general solution of the associated homogeneous equation) and any particular solution ( y_p ) of the non-homogeneous equation. The lecture focuses on non-homogeneous equations with constant coefficients. The three main methods for finding a particular integral are listed: the method of undetermined coefficients (superposition approach), the annihilator operator approach, and the method of variation of parameters.
💡 Why this matters: Understanding that the general solution is ( y = y_c + y_p ) is fundamental to solving all non-homogeneous linear differential equations.
The Method of Undetermined Coefficient
The method is limited to non-homogeneous linear differential equations with constant coefficients.
17.1 The form of Input function ( g(x) )
The input function ( g(x) ) must be of a specific form for the method to be applicable. These forms include:
- A constant function ( k )
- A polynomial function
- An exponential function ( e^{\alpha x} )
- The trigonometric functions ( \sin(\beta x) ) or ( \cos(\beta x) )
- Finite sums and products of these functions.
If ( g(x) ) does not have one of these forms, the method cannot be applied.
17.2 Solution Steps
The solution process involves a systematic series of steps:
- Step 1: Determine the form of the input function ( g(x) ).
- Step 2: Assume the general form of ( y_p ) according to the form of ( g(x) ).
- Step 3: Substitute the assumed ( y_p ) and its derivatives into the given non-homogeneous differential equation.
- Step 4: Simplify and equate coefficients of like terms from both sides of the equation.
- Step 5: Solve the resulting system of equations to find the unknown coefficients.
- Step 6: Substitute the calculated values of the coefficients back into the assumed form for ( y_p ).
17.2.1 Restriction on Input function ( g )
The restriction on the form of ( g(x) ) is necessary because the derivatives of sums and products of polynomials, exponentials, and trigonometric functions are again sums and products of similar kinds of functions. Therefore, the expression ( a y_p'' + b y_p' + c y_p ) must be able to identically equal the input function ( g(x) ).
💡 Why this matters: This restriction ensures that the assumed form for ( y_p ) is an educated guess that can be made to match ( g(x) ).
Caution! A critical point to remember is that the educated guess for ( y_p ) must take into consideration the functions that make up the complementary function ( y_c ). No function in the assumed ( y_p ) must be a solution of the associated homogeneous differential equation. This means the assumed ( y_p ) should not contain terms that duplicate terms in ( y_c ).
🔑 Definition — Complementary function ( y_c ): The general solution of the associated homogeneous linear differential equation. 🔑 Definition — Particular solution ( y_p ): Any solution of the non-homogeneous linear differential equation. 🔑 Definition — Method of Undetermined Coefficients: A technique for finding a particular solution ( y_p ) of a non-homogeneous linear differential equation with constant coefficients, where the input function ( g(x) ) has a specific form (polynomial, exponential, sine/cosine, or sums/products of these). It involves assuming a general form for ( y_p ) with unknown coefficients and then solving for those coefficients.
⭐ Key Takeaways
The method of undetermined coefficients is a direct approach for finding a particular solution to non-homogeneous linear differential equations, but it is restricted to equations with constant coefficients and input functions ( g(x) ) that are constants, polynomials, exponentials, sines, cosines, or sums/products thereof. The core process involves assuming a general form for ( y_p ) that matches the form of ( g(x) ), substituting it into the differential equation, and solving for the unknown coefficients. A crucial caution is that the assumed ( y_p ) must not contain any terms that are already present in the complementary function ( y_c ), as this would lead to an incorrect guess and failure to find a valid solution. This method is a specific case of the superposition approach, where the general solution is constructed as the sum of ( y_c ) and ( y_p ).
🧠 Quick Revision Questions
- What is the general solution of a non-homogeneous linear differential equation?
- List the specific forms that the input function ( g(x) ) must have for the method of undetermined coefficients to be applicable.
- What is the critical caution to consider when assuming the form of the particular solution ( y_p )?
- Describe the six steps involved in the method of undetermined coefficients.
- What does it mean if a term in the assumed ( y_p ) is "duplicated" in the complementary function ( y_c )?
📘 Lecture 18 — Differential Equations (MTH401)
📖 Overview: This lecture introduces the Method of Undetermined Coefficients for solving non-homogeneous linear differential equations with constant coefficients. It explains how to select trial particular solutions based on the form of the input function, including when the input is a sum of functions and when duplication occurs between the particular solution and the complementary function.
🗂️ Topics Covered
This lecture covers trial particular solutions for various input functions including polynomials, exponentials, and trigonometric functions. It discusses how to handle input functions that are sums of terms, how to address duplication between yₚ and y꜀ by multiplying by powers of x, and applies these concepts to second-order and higher-order equations with examples including initial value problems.
📝 Lecture Summary
17.3 Trial particular solutions
This section provides a table of assumed particular solutions yₚ for common input functions g(x). For any constant input, assume yₚ = A. For a linear function (5x + 7), assume yₚ = Ax + B. For a quadratic (3x² − 2), assume yₚ = Ax² + Bx + c. For a cubic (x³ − x + 1), assume yₚ = Ax³ + Bx² + Cx + D. For sin 4x or cos 4x, assume yₚ = A cos 4x + B sin 4x. For e⁵ˣ, assume yₚ = Ae⁵ˣ. For (9x − 2)e⁵ˣ, assume yₚ = (Ax + B)e⁵ˣ. For x²e⁵ˣ, assume yₚ = (Ax² + Bx + C)e⁵ˣ. For e³ˣ sin 4x, assume yₚ = A e³ˣ cos 4x + B e³ˣ sin 4x. For 5x² sin 4x, assume a more complex form involving both cosine and sine terms with quadratic coefficients.
17.4 Input function g(x) as a sum
When the input function g(x) is a sum of m terms of the kind listed in the table, the trial particular solution yₚ is the sum of the trial forms corresponding to each term. This means yₚ is a linear combination of all linearly independent functions generated by repeated differentiation of g(x).
🔑 Definition — Superposition Principle: If g(x) = g₁(x) + g₂(x) + … + gₘ(x) and yₚ₁, yₚ₂, …, yₚₘ are the trial forms, then the particular solution is yₚ = yₚ₁ + yₚ₂ + … + yₚₘ.
📐 Formula: yₚ = yₚ₁ + yₚ₂ + … + yₚₘ (for input function g(x) = g₁(x) + g₂(x) + … + gₘ(x))
📌 Example 1: Solve y″ + 4y′ − 2y = 2x² − 3x + 6 The auxiliary equation m² + 4m − 2 = 0 gives roots m = −2 ± √6, so y꜀ = c₁e^(−2−√6)x + c₂e^(−2+√6)x. For g(x) = 2x² − 3x + 6 (quadratic), assume yₚ = Ax² + Bx + C. After substitution and equating coefficients: −2A = 2, 8A − 2B = −3, 2A + 4B − 2C = 6. Solving gives A = −1, B = −5/2, C = −9, so yₚ = −x² − (5/2)x − 9. General solution: y = y꜀ + yₚ = c₁e^(−2−√6)x + c₂e^(−2+√6)x − x² − (5/2)x − 9.
📌 Example 2: Solve y″ − y′ + y = 2 sin 3x Auxiliary equation m² − m + 1 = 0 gives m = (1 ± i√3)/2, so y꜀ = e^(x/2)[c₁ cos(√3 x/2) + c₂ sin(√3 x/2)]. For g(x) = sin 3x, assume yₚ = A cos 3x + B sin 3x. Substituting: (−8A − 3B) cos 3x + (3A − 8B) sin 3x = 0·cos 3x + 2 sin 3x. Solving −8A − 3B = 0 and 3A − 8B = 2 gives A = 6/73, B = −16/73. General solution: y = e^(x/2)[c₁ cos(√3 x/2) + c₂ sin(√3 x/2)] + (6/73) cos 3x − (16/73) sin 3x.
📌 Example 3: Solve y″ − 2y′ − 3y = 4x − 5 + 6xe²ˣ Auxiliary equation m² − 2m − 3 = 0 gives m = −1, 3, so y꜀ = c₁e⁻ˣ + c₂e³ˣ. g(x) = (4x − 5) + 6xe²ˣ, so assume yₚ₁ = Ax + B for the polynomial part and yₚ₂ = (Cx + D)e²ˣ for the exponential part. After substitution and equating coefficients: −3A = 4, −2A − 3B = −5, −3C = 6, 2C − 3D = 0. Solving gives A = −4/3, B = 23/9, C = −2, D = −4/3.
17.5 Duplication between yₚ and y꜀
If a function in the assumed yₚ is also present in y꜀, then that function is a solution of the associated homogeneous equation, making the assumption incorrect. To resolve this duplication, multiply the duplicate term(s) by xⁿ, where n is the smallest positive integer that eliminates the duplication.
🔑 Definition — Duplication Rule: If yₚᵢ contains terms that duplicate terms in y꜀, multiply yₚᵢ by xⁿ where n is the least positive integer that eliminates the duplication.
💡 Why this matters: Without addressing duplication, the assumed form would lead to a contradiction (like 0 = 8eˣ), making it impossible to find the particular solution.
📌 Example 4: Find particular solution of y″ − 5y′ + 4y = 8eˣ y꜀ = c₁eˣ + c₂e⁴ˣ. Assuming yₚ = Aeˣ leads to 0 = 8eˣ (wrong). Since eˣ duplicates y꜀, multiply by x: yₚ = Axeˣ. Substituting: Axeˣ + 2Aeˣ − 5Axeˣ − 5Aeˣ + 4Axeˣ = 8eˣ, giving −3Aeˣ = 8eˣ, so A = −8/3.
📌 Example 5: Determine form of particular solution for y″ − 8y′ + 25y = 5x³e⁻ˣ − 7e⁻ˣ y꜀ = e⁴ˣ(c₁ cos 3x + c₂ sin 3x). g(x) = (5x³ − 7)e⁻ˣ, assume yₚ = (Ax³ + Bx² + Cx + D)e⁻ˣ. No duplication since y꜀ contains e⁴ˣ terms, not e⁻ˣ.
📌 Example 6: Determine form of particular solution for y″ − y′ + y = 3x² − 5 sin 2x + 7xe⁶ˣ y꜀ = e^(x/2)[c₁ cos(√3 x/2) + c₂ sin(√3 x/2)]. For g₁(x) = 3x², yₚ₁ = Ax² + Bx + C. For g₂(x) = −5 sin 2x, yₚ₂ = D cos 2x + E sin 2x. For g₃(x) = 7xe⁶ˣ, yₚ₃ = (Fx + G)e⁶ˣ. No duplication with y꜀.
📌 Example 7: Find particular solution of y″ − 2y′ + y = eˣ y꜀ = c₁eˣ + c₂xeˣ (repeated root m = 1). Assuming yₚ = Aeˣ duplicates y꜀. Multiply by x: yₚ = Axeˣ still duplicates (xeˣ is in y꜀). Multiply by x²: yₚ = Ax²eˣ has no duplication. Substituting gives yₚ = (1/2)x²eˣ.
📌 Example 8: Solve initial value problem y″ + y = 4x + 10 sin x, y(π) = 0, y′(π) = 2 y꜀ = c₁ cos x + c₂ sin x. For g₁(x) = 4x, yₚ₁ = Ax + B. For g₂(x) = 10 sin x, assuming yₚ₂ = C cos x + D sin x duplicates y꜀. Multiply by x: yₚ₂ = Cx cos x + Dx sin x. Final yₚ = Ax + B + Cx cos x + Dx sin x. Substituting: Ax + B − 2C sin x + 2Dx cos x = 4x + 10 sin x. Solving: B = 0, A = 4, −2C = 10 gives C = −5, 2D = 0 gives D = 0. So yₚ = 4x − 5x cos x. General: y = c₁ cos x + c₂ sin x + 4x − 5x cos x. Using y(π) = 0: c₁(−1) + c₂(0) + 4π − 5π(−1) = 0 → −c₁ + 4π + 5π = 0 → c₁ = 9π. Using y′(π) = 2: −9π sin π + c₂ cos π + 4 + 5π sin π − 5 cos π = 2 → 0 − c₂ + 4 − 0 + 5 = 2 → −c₂ + 9 = 2 → c₂ = 7. Solution: y = 9π cos x + 7 sin x + 4x − 5x cos x.
📌 Example 9: Solve y″ − 6y′ + 9y = 6x² + 2 − 12e³ˣ y꜀ = c₁e³ˣ + c₂xe³ˣ (repeated root m = 3). For g₁(x) = 6x² + 2, yₚ₁ = Ax² + Bx + C. For g₂(x) = −12e³ˣ, yₚ₂ = De³ˣ duplicates y꜀; multiplying by x gives Dxe³ˣ still duplicates; multiplying by x² gives Dx²e³ˣ with no duplication. Final yₚ = Ax² + Bx + C + Dx²e³ˣ. After substitution: 9Ax² + (−12A + 9B)x + (2A − 6B + 9C) + 2De³ˣ = 6x² + 2 − 12e³ˣ. Solving: 9A = 6 → A = 2/3; −12A + 9B = 0 → −8 + 9B = 0 → B = 8/9; 2A − 6B + 9C = 2 → 4/3 − 16/3 + 9C = 2 → −4 + 9C = 2 → C = 2/3; 2D = −12 → D = −6.
📌 Example 10: Solve y‴ + y″ = eˣ cos x Auxiliary equation m³ + m² = 0 → m²(m + 1) = 0 → m = 0, 0, −1, so y꜀ = c₁ + c₂x + c₃e⁻ˣ. For g(x) = eˣ cos x, assume yₚ = Aeˣ cos x + Beˣ sin x. No duplication (y꜀ has e⁻ˣ and constants, not eˣ cos x). After substitution and equating coefficients: (−2A + 4B)eˣ cos x + (−4A − 2B)eˣ sin x = eˣ cos x. Solving −2A + 4B = 1 and −4A − 2B = 0 gives A = −1/10, B = 1/5.
⭐ Key Takeaways
The method of undetermined coefficients requires matching the form of the particular solution to the input function using the provided table of trial forms. When the input function is a sum, the particular solution is the sum of individual trial solutions. Crucially, if any term in the assumed yₚ duplicates a term in y꜀, it must be multiplied by xⁿ until the duplication is eliminated, starting with n = 1 and increasing as needed. For repeated roots, higher powers of x may be required. This method applies to higher-order linear equations with constant coefficients as long as the input function consists of polynomials, exponentials, sines, cosines, or products thereof.
🧠 Quick Revision Questions
- What trial particular solution would you assume for g(x) = 3x² − 5x + 2?
- How do you handle the input function g(x) = 4x + 3e²ˣ + sin x?
- What happens if your assumed yₚ contains a term that is already in y꜀?
- Find the particular solution for y″ − 2y′ + y = eˣ if the complementary function is y꜀ = c₁eˣ + c₂xeˣ.
- What is the form of yₚ for y‴ + y″ = eˣ cos x?
📘 Lecture 19 — Undetermined Coefficient (Annihilator Operator Approach)
📖 Overview: This lecture introduces the annihilator operator approach for finding particular solutions to non-homogeneous linear differential equations with constant coefficients. It presents differential operators as a compact notation and shows how to construct operators that "annihilate" specific functions, enabling a systematic method for determining particular integrals.
🗂️ Topics Covered
The lecture begins by reviewing non-homogeneous differential equations and the need for complementary functions and particular integrals. It then introduces differential operators and notation using D. The concept of annihilator operators is developed, showing how operators like D^n, (D-α)^n, and (D^2-2αD+(α²+β²))^n annihilate polynomial, exponential, and trigonometric functions. Examples demonstrate finding annihilators for various functions and combinations of functions.
📝 Lecture Summary
18 Undetermined Coefficient (Annihilator Operator Approach)
A non-homogeneous linear differential equation of order n has the form aₙ(dⁿy/dxⁿ) + aₙ₋₁(dⁿ⁻¹y/dxⁿ⁻¹) + ... + a₁(dy/dx) + a₀y = g(x). The associated homogeneous equation is aₙ(dⁿy/dxⁿ) + aₙ₋₁(dⁿ⁻¹y/dxⁿ⁻¹) + ... + a₁(dy/dx) + a₀y = 0. To obtain the general solution, we must find: the complementary function y_c (general solution of the associated homogeneous equation) and any particular solution y_p of the non-homogeneous equation. The general solution equals Complementary Function + Particular Integral. In this lecture, we learn to find particular integrals using the concept of differential annihilator operators.
18.1 Differential Operators
The symbol D denotes the differential operator d/dx, so dy/dx = Dy. D is a linear differential operator because D{af(x) + bg(x)} = aDf(x) + bDg(x). Higher order derivatives are expressed as D²y = d²y/dx², D³y = d³y/dx³, and Dⁿy = dⁿy/dxⁿ. A polynomial expression aₙDⁿ + aₙ₋₁Dⁿ⁻¹ + ... + a₁D + a₀ is also a linear differential operator.
🔑 Definition — Differential Operator: The symbol D represents the operation of differentiation with respect to x, so Dy = dy/dx. It is a linear operator.
📐 Formula: Dⁿ(f) = dⁿf/dxⁿ → the nth derivative of f with respect to x
📌 Example: D(e⁴ˣ) = 4e⁴ˣ, D(5x³ - 6x²) = 15x² - 12x, D(cos 2x) = -2 sin 2x
18.2 Differential Equation in Terms of D
Any linear differential equation can be expressed using D notation. For a 2nd order equation ay″ + by′ + cy = g(x), since dy/dx = Dy and d²y/dx² = D²y, we can write (aD² + bD + c)y = g(x). Defining L = aD² + bD + c, the equation becomes L(y) = g(x). L is a second-order linear differential operator with constant coefficients.
Factorization of a differential operator: An nth-order linear differential operator L = aₙDⁿ + aₙ₋₁Dⁿ⁻¹ + ... + a₁D + a₀ with constant coefficients can be factorized whenever the characteristic polynomial can be factorized. The factors of a linear differential operator with constant coefficients commute.
📐 Formula: (D² + 5D + 6) = (D + 2)(D + 3) → factorization treating D as an algebraic quantity
📌 Example: (D² + 5D + 6)y = (D + 2)(D + 3)y = (D + 3)(D + 2)y, since (D+2)(D+3)y = y″ + 5y′ + 6y and (D+3)(D+2)y = y″ + 5y′ + 6y
18.3 Annihilator Operator
Suppose L is a linear differential operator with constant coefficients and y = f(x) is a sufficiently differentiable function. If L(y) = 0, then L is called an annihilator operator of the function f.
🔑 Definition — Annihilator Operator: A linear differential operator L such that L(f) = 0 for a given function f
The differential operator Dⁿ annihilates each of the functions 1, x, x², ..., xⁿ⁻¹. The polynomial function c₀ + c₁x + ... + cₙ₋₁xⁿ⁻¹ can be annihilated by finding an operator that annihilates the highest power of x. The differential operator (D - α)ⁿ annihilates each of the functions e^(αx), xe^(αx), x²e^(αx), ..., xⁿ⁻¹e^(αx). The differential operator (D² - 2αD + (α² + β²))ⁿ is the annihilator operator of the functions e^(αx)cos(βx), xe^(αx)cos(βx), ..., xⁿ⁻¹e^(αx)cos(βx) and e^(αx)sin(βx), xe^(αx)sin(βx), ..., xⁿ⁻¹e^(αx)sin(βx).
💡 Why this matters: The annihilator approach provides a systematic algebraic method for finding particular solutions. Instead of guessing the form of y_p, we apply an annihilator to both sides of the differential equation, converting the non-homogeneous equation into a homogeneous one that can be solved using standard methods.
📌 Example 5: Since Dx = 0, D²x = 0, D³x² = 0, D⁴x³ = 0, the operators D, D², D³, D⁴ are annihilators of constant, x, x², x³ respectively.
📌 Example 6: For y = 1 - 5x² + 8x³, since D⁴x³ = 0, then D⁴ is the annihilator operator.
📌 Example 7: The differential equation (D - α)ⁿy = 0 has general solution y = c₁e^(αx) + c₂xe^(αx) + ... + cₙxⁿ⁻¹e^(αx), so (D - α)ⁿ annihilates these functions.
📌 Example 8a: For f(x) = e^(5x), (D - 5)e^(5x) = 5e⁵ˣ - 5e⁵ˣ = 0, so annihilator is L = D - 5 (α = 5, n = 1).
📌 Example 8b: For g(x) = 4e^(2x) - 6xe^(2x), (D - 2)²(4e^(2x) - 6xe^(2x)) = 0, so annihilator is L = (D - 2)² (α = 2, n = 2).
📌 Example 10: (D² + 2D + 5)e^(-x)cos(2x) = 0 and (D² + 2D + 5)e^(-x)sin(2x) = 0, so D² + 2D + 5 annihilates these functions (α = -1, β = 2, n = 1).
📌 Example 11: (D² + 1)² = D⁴ + 2D² + 1 annihilates cos x, sin x, x cos x, x sin x (α = 0, β = 1, n = 2).
📌 Example 13: For f(x) = 7 - x + 6 sin(3x), D² annihilates y₁(x) = 7 - x, and (D² + 9) annihilates y₂(x) = 6 sin(3x), so D²(D² + 9) annihilates f(x).
📌 Example 14: For f(x) = e^(-3x) + xe^x, (D + 3) annihilates e^(-3x) and (D - 1)² annihilates xe^x, so (D + 3)(D - 1)² annihilates f(x).
Important properties: If L annihilates y₁ and y₂, then L annihilates their linear combination c₁y₁ + c₂y₂. If L₁ annihilates y₁ and L₂ annihilates y₂, but L₁(y₂) ≠ 0 and L₂(y₁) ≠ 0, then the product L₁L₂ annihilates the sum y₁ + y₂. The annihilator of a function is not unique; we seek the operator of lowest possible order.
⭐ Key Takeaways
The D operator notation transforms differential equations into algebraic form, where D represents differentiation. An annihilator operator is a linear differential operator that yields zero when applied to a specific function or class of functions; for polynomials, Dⁿ is the annihilator; for exponentials e^(αx), (D-α)ⁿ is the annihilator; for sine/cosine combinations, (D² - 2αD + (α²+β²))ⁿ is the annihilator. When finding annihilators for sums of functions, we take the product of the individual annihilator operators, always seeking the lowest possible order operator. The annihilator approach provides a systematic, algebraic method for finding particular solutions to non-homogeneous differential equations by converting them into higher-order homogeneous equations.
🧠 Quick Revision Questions
- What is the annihilator operator for the function f(x) = 5x³ - 2x + 7?
- Find the annihilator operator of lowest order for f(x) = e^(4x) + 3xe^(4x).
- What annihilator operator corresponds to the function f(x) = e^(-2x)cos(3x)?
- If L₁ annihilates y₁ and L₂ annihilates y₂, under what condition does L₁L₂ annihilate y₁ + y₂?
- Write the differential equation y″ + 4y′ - 5y = 3x² - 1 in operator notation using D.