基于积分系统研究背景，问询极限求值技巧是否有底层理论

Great question! I’ve thought about this a lot after coming across that closing note in Maxwell Rosenlicht’s Integration in finite terms—G.H. Hardy’s point that integration isn’t just an "art" but a coherent system really resonates, and it makes perfect sense to ask the same about limits.

Short Answer: Yes, there’s a solid systematic basis #

Nearly every "trick" you learn for solving limits ties back directly to the formal definition of limits (the ε-δ framework) and a set of core theorems that build on that definition. These aren’t arbitrary hacks—they’re intentional applications of fundamental calculus rules to rewrite indeterminate or hard-to-evaluate expressions into forms we can compute.

Let’s break down your example: Multiplying by conjugates #

Take the classic trick of multiplying numerator and denominator by a conjugate (like when solving $\lim_{x \to 0} \frac{\sqrt{1+x} - 1}{x}$). Here’s the underlying theory:

• First, we’re dealing with an indeterminate form ($0/0$), which means we can’t directly apply the limit laws for arithmetic operations (since those require the denominator’s limit to be non-zero, or both numerator and denominator to not be $0/0$).
• Multiplying by the conjugate ($\sqrt{1+x} + 1$) uses the algebraic identity $(a-b)(a+b) = a^2 - b^2$ to eliminate the radical in the numerator. This transforms the expression into one where we can cancel the common $x$ term (valid because $x \to 0$ means $x \neq 0$, so we’re not dividing by zero).
• Once we simplify to $\lim_{x \to 0} \frac{1}{\sqrt{1+x} + 1}$, we can use the continuity of square-root functions (and the limit law for quotients, since the denominator’s limit is now $2$, non-zero) to evaluate the limit directly as $1/2$.

Other common tricks and their foundations #

• Factoring to cancel zero factors: Works for $0/0$ forms (e.g., $\lim_{x \to 2} \frac{x^2 - 4}{x-2}$) because we’re using polynomial factorization to rewrite the expression into a form where the limit laws apply—again, relying on the fact that $x \neq 2$ as $x \to 2$, so cancellation is valid.
• Equivalent infinitesimal substitution: Rooted in Taylor series expansions and the definition of asymptotic equivalence ($f(x) \sim g(x)$ as $x \to a$ means $\lim_{x \to a} \frac{f(x)}{g(x)} = 1$). This lets us replace complicated terms with simpler ones while preserving the limit value.
• Squeeze Theorem: Directly derived from the ε-δ definition of limits—it formalizes the idea that if a function is trapped between two functions with the same limit, it must share that limit.

Why it feels like an "art" sometimes #

Even with this systematic foundation, applying these rules requires judgment: knowing which trick to use when, how to rewrite an expression, or recognizing a pattern. But this is just applying a structured toolkit, not creating art out of thin air. Hardy’s point about integration applies here too—what seems like arbitrary skill is actually mastery of a coherent, rule-based system.

内容的提问来源于stack exchange，提问作者John Mack