📚 Difference Equations: Theory, Methods and IB Applications | 差分方程：理论、方法与IB应用
Difference equations describe the evolution of a sequence where each term is defined in terms of previous terms. They appear in many areas of mathematics, including modelling discrete dynamical systems, finance, and population biology. In the IB Diploma Programme, both Analysis and Approaches (AA) and Applications and Interpretation (AI) cover recurrence relations and their solutions, making difference equations a key topic for higher level students.
差分方程描述数列的演化，其中每一项都由前面的项定义。它们出现在许多数学领域中，包括离散动力系统建模、金融和种群生物学。在IB文凭课程中，分析与方法（AA）以及应用与解释（AI）都涵盖递推关系及其解法，使得差分方程成为高水平学生的一个重要主题。
1. Understanding Difference Equations | 理解差分方程
the general form yₙ₊₁ = f(yₙ, n). The order of the equation is the difference between the highest and the lowest subscripts that appear. In an IB context, difference equations usually appear as linear recurrence relations with constant coefficients. For example, yₙ₊₁ = 2yₙ + 1 defines a first-order linear non-homogeneous difference equation: the term 2yₙ is the homogeneous part and the constant 1 is the non-homogeneous part.
差分方程将一个数列的某一项与前面一项或几项联系起来。对于数列 y₀, y₁, y₂, ...，一阶差分方程的一般形式为 yₙ₊₁ = f(yₙ, n)。方程的阶数，等于方程中出现的最高下标与最低下标之差。在 IB 情境中，差分方程通常表现为常系数线性递推关系。例如 yₙ₊₁ = 2yₙ + 1 定义了一个一阶线性非齐次差分方程：其中 2yₙ 是齐次部分，常数 1 是非齐次部分。
the equation Δyₙ = 2yₙ + 1 is exactly the same statement as yₙ₊₁ = 3yₙ + 1 after rearrangement.
最基本的构件是差分算子 Δ，定义为 Δyₙ = yₙ₊₁ − yₙ。高阶差分通过反复作用 Δ 得到，因此 Δyₙ = 2yₙ + 1 与整理后的 yₙ₊₁ = 3yₙ + 1 是完全等价的表述。
yₙ₊₁ = a yₙ + b ⇔ yₙ₊₁ − yₙ = (a − 1)yₙ + b
2. Discrete Versus Continuous Models | 离散模型与连续模型
ered steps. A differential equation such as dy/dx = ky states how fast a quantity changes at every instant, while the difference equation yₙ₊₁ − yₙ = kyₙ states how much the quantity changes from one step to the next. The two are linked by the approximation dy/dx ≈ (yₙ₊₁ − yₙ)/h for a step size h, which is why a difference equation can be read as a numerical method for a differential equation.
与由微分方程建模的连续系统不同，差分方程是离散的：变量只在整数步上取值。像 dy/dx = ky 这样的微分方程描述的是某一量在每一瞬间的变化速度，而差分方程 yₙ₊₁ − yₙ = kyₙ 描述的是该量从一步到下一步的变化量。二者通过近似 dy/dx ≈ (yₙ₊₁ − yₙ)/h 相联系（h 为步长），这正是差分方程可以被看作微分方程数值解法的原因。
Discrete models are the natural choice whenever data arrive in generations, periods or cycles: interest is credited once a year, a population breeds once a season, a patient takes a tablet once every eight hours. A useful exam habit is to ask, for every problem, whether the question is describing a rate at an instant (differential equation) or a jump between two labelled states (difference equation).
当数据按“代”“期”或“周期”出现时，离散模型是自然的选择：利息每年结算一次，种群每季繁殖一次，病人每八小时服一次药。一个有用的考试习惯是：对每一道题都先判断它描述的是某一瞬间的变化率（微分方程），还是两个被标记状态之间的跳跃（差分方程）。
3. Where the Topic Appears in the IB | IB 课程中的定位
h a given first term, and students are expected to generate terms, describe long-term behaviour and solve first-order linear recurrences. In Analysis and Approaches the emphasis falls on algebraic technique and proof: finding closed forms, justifying convergence with |a| on the emphasis falls on modelling and interpretation: choosing a sensible recurrence, using technology to iterate it, and explaining what the fixed point or the growth rate means in context.
两门课程都会较早引入递推关系，通常以 uₙ₊₁ = 3uₙ − 2 这类给定首项的数列出现，学生需要生成项、描述长期行为并求解一阶线性递推。在“分析与方法”中，重点落在代数技巧与论证上：求通解、用 |a| < 1 论证收敛、用特征方程求解二阶递推。在“应用与解释”中，重点落在建模与解释上：选择合适的递推关系、用计算器迭代、并说明不动点或增长率在情境中的含义。
For both routes the same core skills are examined: writing a recurrence from a word problem, iterating accurately, finding an equilibrium value, determining whether the sequence converges, and using an initial condition to fix the constants in a general solution. Higher level students additionally meet systems of recurrences and matrix iteration, which is where the topic connects to linear algebra.
无论哪条路径，考查的核心能力相同：由文字情境写出递推式、准确迭代、求平衡值、判断数列是否收敛，以及利用初始条件确定通解中的常数。高水平学生还会遇到递推方程组与矩阵迭代，这正是本主题与线性代数衔接的地方。
4. First-Order Linear Recurrences | 一阶线性递推
ng y* = b/(1 − a) provided a ≠ 1. Shifting the sequence by the equilibrium, vₙ = yₙ − y*, turns the non-homogeneous recurrence into the homogeneous one vₙ₊₁ = a vₙ, so vₙ = aⁿ v₀ and therefore:
yₙ₊₁ = a yₙ + b（a、b 为常数）是最重要的情形。令 yₙ₊₁ = yₙ = y* 可得平衡点（不动点）y* = b/(1 − a)，前提是 a ≠ 1。把数列平移一个平衡量，令 vₙ = yₙ − y*，非齐次递推就化为齐次递推 vₙ₊₁ = a vₙ，从而 vₙ = aⁿ v₀，于是有：
yₙ = aⁿ (y₀ − b/(1 − a)) + b/(1 − a), a ≠ 1
This single formula answers almost every one-line convergence question in the syllabus, because convergence depends only on the multiplier a.
这一个公式几乎能回答大纲中所有一行式的收敛问题，因为收敛性只取决于乘数 a。
| Value of a | a 的取值 | Behaviour as n → ∞ | n → ∞ 时的行为 | Convergent? | 是否收敛 |
| a > 1 | Monotone away from equilibrium | 单调远离平衡点 | No | 否 |
| 0 < a < 1 | Monotone towards equilibrium | 单调趋向平衡点 | Yes, to b/(1 − a) | 是，收敛到 b/(1 − a) |
| −1 < a < 0 | Oscillating decay | 振荡衰减 | Yes | 是 |
| a < −1 | Oscillation with growing amplitude | 振幅增大的振荡 | No | 否 |
| a = 1 | Arithmetic sequence, step b | 公差为 b 的等差数列 | No (unless b = 0) | 否（除非 b = 0） |
| a = −1 | Period-2 cycle | 周期为 2 的循环 | No, but bounded | 否，但有界 |
5. Equilibrium Points and Stability | 平衡点与稳定性
point satisfies y* = f(y*). Stability is decided by the derivative of f at that point: if |f′(y*)| < 1 the equilibrium is stable (attracting), if |f′(y*)| > 1 it is unstable (repelling), and if |f′(y*)| id to be neutral. For the linear case f(y) = ay + b this reduces to the familiar condition |a| < 1.
对于一般递推 yₙ₊₁ = f(yₙ)，平衡点满足 y* = f(y*)。稳定性由 f 在该点的导数决定：若 |f′(y*)| < 1，平衡点是稳定的（吸引的）；若 |f′(y*)| > 1，则是不稳定的（排斥的）；若 |f′(y*)| = 1，判别失效，称为中性。在线性情形 f(y) = ay + b 下，这就退化为熟悉的 |a| < 1。
from y₀ on the horizontal axis and move vertically to the curve, then horizontally to the line, and repeat. If the path spirals inward towards the intersection the equilibrium is stable; if it spirals outward it is unstable. This is one of the fastest ways to answer an AI-style modelling question about long-term behaviour.
在图像上，稳定性可以通过蛛网图或阶梯图看出：同时画出 y = f(x) 与直线 y = x，从横轴上的 y₀ 出发先竖直走到曲线上，再水平走到直线上，重复此过程。若路径向内盘旋靠近交点，平衡点稳定；若向外盘旋，则不�稳定。这是回答 AI 风格长期行为建模题最快的方法之一。
6. Second-Order Linear Recurrences | 二阶线性递推
= λⁿ gives λⁿ⁺² = p λⁿ⁺¹ + q λⁿ, and dividing by λⁿ produces the characteristic equation:
二阶线性递推的形式为 yₙ₊₂ = p yₙ₊₁ + q yₙ。代入试探解 yₙ = λⁿ 得 λⁿ⁺² = p λⁿ⁺¹ + q λⁿ，两边除以 λⁿ，即得特征方程：
λ² − p λ − q = 0
The three cases of the discriminant determine the shape of the general solution:
判别式的三种情形决定了通解的形式：
- Distinct real roots λ₁ ≠ λ₂: yₙ = A λ₁ⁿ + B λ₂ⁿ, with A and B fixed by two initial conditions. | 两个不同实根 λ₁ ≠ λ₂：yₙ = A λ₁ⁿ + B λ₂ⁿ，其中 A、B 由两个初始条件确定。
- Repeated root λ: yₙ = (A + Bn) λⁿ.
广告 / Sponsored现货 Cambridge Biology for the IB Diploma coursebook 剑桥国际文凭 I…生物 备考教辅 · 正版现货 · 与本文内容相关查看详情 / View on Taobao参考价 ¥546.0
更多咨询请联系16621398022（同微信）
备考资讯每日更新，关注不错过考纲变化 · 科目分析 · 学习方法
在微信里长按二维码识别关注
或截图后用「扫一扫 → 相册」选择本图公众号：tutorhao