Introduction to Ordinary Differential Equations

Understanding Ordinary Differential Equations What is an Ordinary Differential Equation? An ordinary differential equation (or ODE) is an equation that relates an unknown function and its derivatives. The key word "ordinary" means the unknown function depends on only one independent variable, distinguishing these equations from partial differential equations, which involve multiple independent variables. The general form of an ordinary differential equation is: $$F\bigl(x,\,y(x),\,y'(x),\,y''(x),\dots ,y^{(n)}(x)\bigr)=0$$ Here, $x$ is the independent variable (think of it as the "input"), and $y(x)$ is the dependent variable we're trying to find (the "output"). The symbols $y'$, $y''$, and $y^{(n)}$ represent the first, second, and $n$-th derivatives of $y$ with respect to $x$. The order of an ODE is determined by the highest derivative that appears in the equation. An equation with only $y'$ is first-order, one with $y''$ is second-order, and so on. Classifying Equations: Order and Linearity Two fundamental properties help us classify ODEs and determine which solution technique to use: Order tells us the highest derivative present. Simply look at the derivatives in your equation and identify the highest one. Linearity describes how the dependent variable and its derivatives appear. An equation is linear if the unknown function $y$ and all its derivatives appear only to the first power and are never multiplied together. A linear ODE has the general form: $$an(x)\,y^{(n)}+a{n-1}(x)\,y^{(n-1)}+\dots +a1(x)\,y' + a0(x)\,y = g(x)$$ For example, $y'' + 3y' + 2y = 0$ is linear (all terms with $y$, $y'$, and $y''$ are to the first power), but $y' + y^2 = 0$ is not linear because $y$ is squared. This distinction matters because we have different methods for solving linear versus nonlinear equations. First-Order Separable Equations One of the most important classes of ODEs you'll solve is the separable first-order equation. An equation is separable if it can be written as: $$\frac{dy}{dx}=g(x)\,h(y)$$ The name "separable" comes from the key technique: we can rearrange to separate the variables onto opposite sides: $$\frac{dy}{h(y)}=g(x)\,dx$$ Once separated, we integrate each side independently: $$\int \frac{dy}{h(y)} = \int g(x)\,dx$$ Why this works: When you see $\frac{dy}{dx}$, remember it's a ratio of differentials. You can manipulate it algebraically just like any fraction—multiply both sides by $h(y)$ and by $dx$ to move all $y$ terms to one side and all $x$ terms to the other. After integrating both sides, solve for $y$ to get your solution. This technique is straightforward once you recognize the separable form, making it a powerful tool for a common class of first-order equations. First-Order Linear Equations A first-order linear equation has the specific form: $$y'+p(x)\,y = q(x)$$ where $p(x)$ and $q(x)$ are known functions of $x$ alone. The solution strategy uses an integrating factor, a carefully chosen function that transforms the left side into a perfect derivative. The integrating factor is: $$\mu(x)=e^{\int p(x)\,dx}$$ Multiply both sides of the equation by $\mu(x)$: $$\mu(x)\,y'+\mu(x)\,p(x)\,y = \mu(x)\,q(x)$$ The magic is that the left side is now the derivative of a product: $\frac{d}{dx}\bigl[\mu(x)\,y\bigr] = \mu(x)\,q(x)$. You can then integrate both sides to solve for $y$. This method works because the integrating factor is specifically constructed so that $\frac{d\mu}{dx} = \mu(x)\,p(x)$, which makes the left side simplify perfectly. Higher-Order Linear Equations with Constant Coefficients For higher-order linear equations with constant coefficients, we have a systematic approach. These equations have the form: $$an\,y^{(n)}+a{n-1}\,y^{(n-1)}+\dots +a1\,y' + a0\,y = g(x)$$ where all the $ai$ are constants (not functions of $x$). The Characteristic Polynomial Method When $g(x) = 0$ (the homogeneous case), we propose a trial solution $y=e^{\lambda x}$ for some constant $\lambda$ to be determined. Substituting $y=e^{\lambda x}$ into the equation gives us a polynomial equation in $\lambda$ called the characteristic polynomial. Solving this polynomial yields roots $\lambda1, \lambda2, \ldots$. Each root contributes to the homogeneous solution as follows: For each real root $\lambda$: contribute a term $e^{\lambda x}$ to the solution For each pair of complex conjugate roots $\alpha \pm i\beta$: contribute terms $e^{\alpha x}\cos(\beta x)$ and $e^{\alpha x}\sin(\beta x)$ to the solution The general homogeneous solution is a linear combination of all these contributions: $$yh(x) = c1\,(\text{first solution}) + c2\,(\text{second solution}) + \cdots$$ where $c1, c2, \ldots$ are arbitrary constants determined by initial conditions. Handling the Non-Homogeneous Case For the full equation (when $g(x) \ne 0$), find any particular solution $yp(x)$ that satisfies the non-homogeneous equation. The general solution is then: $$y(x) = yh(x) + yp(x)$$ This is the principle of superposition: combine the homogeneous solution with any particular solution to get the full picture. Initial-Value Problems An initial-value problem pairs a differential equation with initial conditions that specify the value of $y$ and possibly its derivatives at a particular point. For example: First-order: $y' = f(x, y)$ with $y(x0) = y0$ Second-order: $y'' = f(x, y, y')$ with $y(x0) = y0$ and $y'(x0) = y1$ The initial conditions serve a crucial purpose: they pin down the arbitrary constants in the general solution. When you solve a differential equation, you typically get a family of solutions with multiple arbitrary constants. The initial conditions select exactly one member of this family—the unique solution that satisfies both the equation and the specified initial data. Existence and Uniqueness For first-order linear equations and separable equations with continuous right-hand sides, there is a uniqueness theorem: a unique solution exists and passes through any specified initial point. This gives us confidence that our initial-value problem has exactly one answer—no more, no fewer. This guarantee relies on mild assumptions (continuity, smoothness), which are almost always satisfied in practice. Understanding that solutions are unique helps you recognize when you've found the correct answer to an initial-value problem.