Foundations, Classifications & Geometric Theory of Differential Equations

1. Fundamental Definitions & Terminology

A differential equation (DE) is any mathematical equation involving an unknown function $y = \phi(x)$ and one or more of its derivatives with respect to one or more independent variables. If the unknown function depends on only one independent variable, the equation is called an Ordinary Differential Equation (ODE). If the unknown function depends on two or more independent variables, the equation involves partial derivatives and is designated a Partial Differential Equation (PDE).

The general implicit representation of an $n$-th order ODE in a single dependent variable $y$ and independent variable $x$ is expressed as:

2. Order and Degree of an Ordinary Differential Equation

Two foundational topological and algebraic integers characterize any ODE:

• Order: The order of a differential equation is the order of the highest derivative appearing in the equation. For instance, $\frac{d^2y}{dx^2} + 5\left(\frac{dy}{dx}\right)^3 + y = 0$ is of order 2.
• Degree: The degree of a differential equation is the power (algebraic exponent) to which the highest derivative is raised, after the equation has been rationalized and cleared of all fractional powers and radicals with respect to the derivatives. For example, in $\left[1 + \left(\frac{dy}{dx}\right)^2\right]^{3/2} = k \frac{d^2y}{dx^2}$, squaring both sides yields $\left[1 + (y')^2\right]^3 = k^2 (y'')^2$, confirming that the equation has order 2 and degree 2.

3. The Criterion of Linearity

An $n$-th order ordinary differential equation is classified as linear if it can be written in the form:

Linearly characterized equations must strictly satisfy two defining properties:

1. The dependent variable $y$ and all its derivatives $y', y'', \dots, y^{(n)}$ appear strictly to the first power (no terms like $y^2, (y')^3, \sqrt{y'}$).
2. No products of the dependent variable and its derivatives appear (no terms like $y \cdot y'$ or $y' \cdot y''$).
3. Coefficients $a_k(x)$ and the forcing term $g(x)$ depend solely on the independent variable $x$.

If any of these conditions are violated, the equation is non-linear (e.g., $y'' + \sin(y) = 0$ is non-linear due to the transcendental term in $y$; $y y' + x = 0$ is non-linear due to the product $y y'$).

1. Formation of Differential Equations by Eliminating Arbitrary Constants

In physical modeling and analytical geometry, a family of curves is defined by an algebraic equation containing $n$ arbitrary parameters (constants) $C_1, C_2, \dots, C_n$:

To eliminate these $n$ arbitrary constants, we differentiate the relation successively $n$ times with respect to $x$, yielding a system of $n + 1$ equations involving $x, y, y', y'', \dots, y^{(n)}$ and $C_1, \dots, C_n$. Eliminating the $n$ constants among these $n+1$ relations produces an ODE of order $n$.

Fundamental Theorem: The elimination of $n$ independent arbitrary constants from a relation yields a differential equation of order exactly $n$.

2. Taxonomies of Solutions: General, Particular & Singular

• General Solution: An explicit or implicit relation $\phi(x, y, C_1, \dots, C_n) = 0$ that satisfies the $n$-th order ODE and contains exactly $n$ essential arbitrary constants.
• Particular Solution: Any solution obtained directly from the general solution by assigning specific numerical values to one or more of the arbitrary constants $C_k$, typically dictated by initial or boundary constraints.
• Singular Solution: A solution that cannot be obtained from the general solution by any choice of the arbitrary constants. Geometrically, singular solutions represent envelopes of the family of curves represented by the general solution.

3. Initial Value Problems (IVPs) versus Boundary Value Problems (BVPs)

An Initial Value Problem (IVP) prescribes conditions on the unknown function and its derivatives at a single value of the independent variable $x_0$:

A Boundary Value Problem (BVP), by contrast, prescribes conditions at two or more distinct points (e.g., $y(a) = \alpha$, $y(b) = \beta$). While IVPs generally possess unique solutions under Lipschitz conditions, BVPs can have a unique solution, infinitely many solutions, or no solution at all.

1. Statement of the Picard-Lindelöf Theorem

If $f(x, y)$ is continuous on $R$ and satisfies a Lipschitz condition with respect to $y$ in $R$, that is, there exists a constant $L > 0$ such that:

then there exists a unique solution $y = \phi(x)$ to the initial value problem $y' = f(x, y), y(x_0) = y_0$, defined on an interval $|x - x_0| \le h$, where $h = \min\left(a, \frac{b}{M}\right)$ and $M = \max_{(x,y) \in R} |f(x, y)|$.

2. Picard's Method of Successive Approximations

Integrating $y' = f(t, y(t))$ from $x_0$ to $x$ converts the differential initial value problem into an equivalent Volterra integral equation:

Picard's iterative algorithm constructs a sequence of continuous approximations $\{\phi_k(x)\}_{k=0}^\infty$ defined recursively by:

By Banach's fixed-point theorem on the complete metric space $C([x_0-h, x_0+h])$, this sequence converges uniformly to the unique continuous solution $\phi(x) = \lim_{k\to\infty} \phi_k(x)$.

3. Step-by-Step Analytical Example

Consider the IVP: $y' = 2x(1 + y)$ with $y(0) = 0$. Here $x_0 = 0, y_0 = 0$, and $f(x, y) = 2x(1 + y)$:

Direct differentiation confirms: $\phi'(x) = 2x e^{x^2} = 2x(1 + (e^{x^2}-1)) = 2x(1 + y)$ with $\phi(0) = e^0 - 1 = 0$, yielding the exact closed-form solution!

1. Direction Fields (Slope Fields) & Geometric Interpretation

Even when a first-order ODE $y' = f(x, y)$ cannot be solved by elementary analytical integration, the equation provides immediate geometric insight: at every point $(x, y)$ in the plane where $f(x, y)$ is defined, the derivative $y'$ represents the slope of the tangent line to the integral curve passing through that point. A direction field is a graphical grid of small line segments possessing slope $f(x, y)$.

2. The Method of Isoclines

An isocline is a curve along which the slopes of the integral curves are constant. Setting $f(x, y) = c$, where $c$ is a constant parameter, yields the family of isoclines. Integral curves cross each isocline $f(x, y) = c$ with precisely the slope $c$. Sketching several isoclines provides an exact scaffolding for tracing solution trajectories.

3. Autonomous Equations & Phase Line Stability

An ODE is called autonomous if the independent variable $x$ (often time $t$) does not appear explicitly:

The zeros of $f(y)$, where $f(c) = 0$, are called equilibrium points (or critical points). The constant functions $y(t) \equiv c$ are equilibrium solutions. The qualitative behavior of all non-equilibrium solutions is classified on the one-dimensional phase line:

• Asymptotically Stable (Attractor / Sink): If $f'(c) < 0$, trajectories on both sides move toward $c$ as $t \to \infty$.
• Unstable (Repeller / Source): If $f'(c) > 0$, trajectories on both sides move away from $c$ as $t \to \infty$.
• Semi-stable (Shunt): Trajectories approach $c$ from one side and diverge on the other (occurs when $f(y)$ does not change sign across $c$, such as when $f(y)$ has a double root).