📚 Exponential Modelling | 指数建模
ling is a core topic in Edexcel A-Level Mathematics. It uses functions of the form y = aeᵏˣ or y = abˣ criticise these models is essential for both pure mathematics papers and applied contexts.
指数建模是爱德思 A-Level 数学的核心内容。它使用形式为 y = aeᵏˣ 或 y = abˣ 的函数，来描述以与当前值成比例的速率增长或衰减的量。掌握如何建立、解释、求解和评价这些模型，对纯数学试卷和实际应用情境都至关重要。
1. The Idea of Exponential Change | 指数变化的概念
ontinuous compound interest all follow this pattern. In mathematical terms, if dy/dx = ky, then y is an exponential function of x, and the constant k is called the rate constant. The larger the magnitude of k, the faster the quantity changes.
当量的变化率与量本身成正比时，就出现指数变化。人口增长、放射性衰变和连续复利都遵循这一规律。用数学语言来说，如果 dy/dx = ky，那么 y 就是 x 的指数函数，常数 k 称为速率常数。k 的绝对值越大，量的变化就越快。
dy/dx = ky ⇒ y = aeᵏˣ
A useful way to recognise exponential behaviour in data is to look at ratios rather than differences. If a quantity is multiplied by the same factor in every equal time interval, the data are exponential; if it increases by the same amount, they are linear.
识别数据是否呈指数变化的一个实用方法是看比值而不是差值。如果量在每一个相等时间间隔内都乘以同一个倍数，数据就是指数的；如果每次增加相同的量，数据就是线性的。
2. Standard Exponential Form and Parameters | 标准指数形式与参数
The most common model in Edexcel questions is y = aeᵏˣ y grows or decays, and the sign of k determines whether the model represents growth or decay.
爱德思题目中最常见的模型是 y = aeᵏˣ，其中 a 是 x = 0 时的初始值，k 是速率常数。由于 e⁰ = 1，代入 x = 0 可得 y = a。k 的值控制量增长或衰减的快慢，而 k 的符号决定模型代表增长还是衰减。
y = aeᵏˣ, a = y(0), k = rate constant
In context questions you are usually expected to state the meaning of both parameters. For example, if y is the number of bacteria in a culture and x is time in hours, then a is the initial number of bacteria and k is the continuous growth rate per hour.
在情境题中，通常需要说明两个参数的实际含义。例如，若 y 是培养皿中细菌的数量、x 是时间（小时），那么 a 是初始细菌数，k 是每小时的连续增长率。
3. Growth and Decay Conditions | 增长与衰减条件
For y = aeᵏˣ ble below summarises the two cases, and it is worth memorising the shape of each graph because sketch questions appear regularly.
对于 a > 0 的模型 y = aeᵏˣ，如果 k 为正，y 随 x 增大而无界增长；如果 k 为负，y 逐渐趋近于零，但不会变成负数。下表总结了这两种情况。由于画图题经常出现，记住两种图形的形状很有必要。
| Condition | Behaviour |
| k > 0 | Growth: y increases as x increases (增长：y 随 x 增大而增大) |
| k < 0 | Decay: y decreases towards 0 (衰减：y 趋近于 0) |
| k = 0 | Constant: y = a for all x (恒定：y 恒等于 a) |
asymptote. The curve always passes through (0, a), always lies above the x-axis when a > 0, and is either strictly increasing or strictly decreasing, so it has no turning points.
对于衰减曲线，y = 0 是水平渐近线。曲线总经过点 (0, a)；当 a > 0 时曲线恒在 x 轴上方；曲线严格递增或严格递减，因此没有极值点。
4. Base e and Base b Forms | 以 e 为底与以 b 为底的形式
Some problems use the base-b form y = abˣ he form that matches the data: if the ratio between consecutive values is constant, y = abˣ ler; if a rate is given as a percentage per unit time, the e-form is usually more natural.
有些问题使用以 b 为底的模型 y = abˣ。由于 b = eᵏ，这两种形式是等价的，取自然对数可得 k = ln b。应根据数据选择合适的形式：如果相邻值的比值恒定，y = abˣ 通常更简便；如果给定的是单位时间的百分比速率，则 e 形式通常更自然。
b = eᵏ ⇔ k = ln b
e multiplier per unit increase in x. A value of b = 1.05 means the quantity increases by 5% each time x increases by 1, whereas b = 0.8 means it falls by 20%.
注意，在以 b 为底的形式中，参数 b 有清晰的解释：它是 x 每增加一个单位时的乘数。b = 1.05 表示 x 每增加 1，量增加 5%；而 b = 0.8 表示量下降 20%。
5. Linearising with Logarithms | 用对数线性化
ta can be transformed into linear data by taking natural logarithms. Starting with y = aeᵏˣ to test whether data follow an exponential trend and to estimate parameters from a graph.
指数数据可以通过取自然对数转换为线性数据。从 y = aeᵏˣ 出发，两边取自然对数得 ln y = ln a + kx。这是形如 Y = mx + c 的直线关系，其中 Y = ln y、斜率 m = k、截距 c = ln a。它可用于检验数据是否呈指数趋势，并从图象估计参数。
y = aeᵏˣ ⇒ ln y = ln a + kx (plot ln y against x)
The same idea works for other bases. For y = abˣ s is strong evidence for an exponential model, and the gradient gives log b so that b = 10gradient when base-10 logs are used.
同样的思路适用于其他底数。对于 y = abˣ，取对数得 log y = log a + (log b)x。若以 log y 对 x 作图，点近似落在一条直线上，这就是指数模型的有力证据；斜率给出 log b，因此使用以 10 为底的对数时 b = 10斜率。
6. Finding a and k from Two Data Points | 由两个数据点求 a 和 k
If you are told two pairs of values (x₁, y₁) and (x₂, y₂) satisfy y = aeᵏˣ ln a + kx₂. Subtracting eliminates ln a and gives an exact formula for the rate constant.
若已知两组数据 (x₁, y₁) 和 (x₂, y₂) 满足 y = aeᵏˣ，对两式取对数得 ln y₁ = ln a + kx₁ 与 ln y₂ = ln a + kx₂。两式相减可消去 ln a，得到速率常数的精确公式。
k = (ln y₂ − ln y₁) / (x₂ − x₁) = ln(y₂ ÷ y₁) / (x₂ − x₁)
Once k is known, substitute either point back to find a: a = y₁ e−kx₁ which is often the quickest route in an exam because it avoids handling logarithms twice.
求出 k 后，把任一点代回即可求 a：a = y₁ e−kx₁。另一种做法是把两式相除得 y₂ ÷ y₁ = e^(k(x₂ − x₁))，这往往是考试中最快的路径，因为它避免重复处理对数。
Always keep full calculator precision until the final line. Rounding k to two decimal places in the middle of a multi-part question is a very common cause of lost accuracy marks.
在最后一步之前，应保留计算器的全部精度。在多小问题目中把 k 提前四舍五入到两位小数，是丢失精度分的常见原因。
7. Percentage Rates, Doubling Time and Half-Life | 百分比速率、倍增时间与半衰期
When a quantity grows by r% is b = 1 + r ÷ 100, so the continuous rate constant is k = ln(1 + r ÷ 100). For example, 8% growth per year corresponds to k = ln 1.08 ≈ 0.0770, which is slightly less than 0.08.
当量每单位时间增长 r% 时，离散乘数为 b = 1 + r ÷ 100，因此连续速率常数为 k = ln(1 + r ÷ 100)。例如每年增长 8% 对应 k = ln 1.08 ≈ 0.0770，略小于 0.08。
k = ln(1 + r ÷ 100), doubling time T = ln 2 ÷ k, half-life T = ln 2 ÷ |k|
Doubling time and half-life follow from setting y = 2a or y = a ÷ 2. For growth, aeᵏᵀ = 2a gives T = ln 2 ÷ k; for decay, ae−|k|T = a ÷ 2 gives T = ln 2 ÷ |k| ame value of T applies at every stage, which is the defining feature of exponential decay.
倍增时间与半衰期可由令 y = 2a 或 y = a ÷ 2 得到。增长情形下，aeᵏᵀ = 2a 给出 T = ln 2 ÷ k；衰减情形下，ae−|k|T = a ÷ 2 给出 T = ln 2 ÷ |k|。T 在任何阶段都相同，这正是指数衰减的本质特征。
8. Solving Exponential Equations and Inequalities | 求解指数方程与不等式
To solve aeᵏˣ = c, first isolate the exponential: eᵏˣ k) ln(c ÷ a). The method requires c ÷ a > 0, otherwise there is no solution because eᵏˣ is always positive.
解 aeᵏˣ = c 时，先分离指数项：eᵏˣ = c ÷ a，再两边取自然对数并除以 k，得 x = (1 ÷ k) ln(c ÷ a)。该方法要求 c ÷ a > 0；否则无解，因为 eᵏˣ 恒为正。
aeᵏˣ = c ⇒ x = (1 ÷ k) ln(c ÷ a)
For y = abˣ ction of the inequality sign, so it is safest to solve the corresponding equation first and then test a value.
对于 y = abˣ，同一方程可用以 b 为底的对数求解：x = log(c ÷ a) ÷ log b。解不等式时要注意，除以负的 k 会改变不等号方向，因此最稳妥的做法是先解对应方程，再代入一个值检验。
9. Differentiation and Integration of Exponential Models | 指数模型的微分与积分
Differentiating y = aeᵏˣ gives dy/dx = kaeᵏˣ ky and the solution form, and it is frequently examined in both pure and mechanics-style contexts.
对 y = aeᵏˣ 求导得 dy/dx = kaeᵏˣ = ky，这验证了变化率与当前值成正比。该结果正是建模假设 dy/dx = ky 与解的形式之间的桥梁，在纯数学与应用情境中都经常考查。
y = aeᵏˣ ⇒ dy/dx = ky; ∫ eᵏˣ dx = (1 ÷ k) eᵏˣ + C
Integration reverses the process: ∫ eᵏˣ dx = (1 ÷ k)eᵏˣ antity accumulated over a time interval when the rate is given as an exponential function.
积分是其逆过程：∫ eᵏˣ dx = (1 ÷ k)eᵏˣ + C（k ≠ 0）。定积分则用原函数在上下限处求值。应用包括当速率以指数函数给出时，求某时间段内累积的总量。
10. Criticising and Validating Models | 评价与检验模型
Every exponential model rests on assumptions, and examiners reward candidates who state them clearly. The most important assumption is that k remains constant, which is rarely true for long periods; real populations are limited by food, space and disease, and radioactive decay changes only when the substance itself changes.
每个指数模型都建立在假设之上，清楚说明这些假设会得到考官的认可。最重要的假设是 k 保持不变，而在长时间尺度上这很少成立；真实种群受食物、空间和疾病限制，而放射性衰变只有在物质本身改变时才改变。
Other limitations worth mentioning are that the model is continuous while data may be recorded at discrete intervals, that extrapolating far beyond the data range is unreliable, and that small percentage errors in k are hugely amplified when the model is used to predict large x values.
其他值得指出的局限包括：模型是连续的，而数据可能是按离散间隔记录的；远超数据范围的外推不可靠；当模型用于预测很大的 x 值时，k 中很小的百分比误差会被极大地放大。
Check: constant ratio, straight-line ln y against x, sensible long-term behaviour
A good validation routine is: convert the data to log form and check for linearity, comment on the value of the correlation coefficient or on residuals, and then judge whether the prediction is physically possible in context.
良好的检验流程是：把数据转换为对数形式并检查线性，说明相关系数或残差的情况，然后判断预测在情境中是否物理上可行。
11. Exam Technique and Common Errors | 考试技巧与常见错误
Most marks in this topic are lost through careless algebra rather than misunderstanding. Always write the model down before substituting, keep exact values such as ln 1.08 until the end, and state the units of your answer whenever the question gives a context.
本专题失分大多源于粗心的代数运算，而非理解错误。应先把模型写出来再代入；像 ln 1.08 这样的精确值要保留到最后；题目给情境时，答案要写出单位。
- Do not write ln(a eᵏˣ) = kx; the correct expansion is ln a + kx. (不要把 ln(a eᵏˣ) 写成 kx，正确展开是 ln a + kx。)
- When plotting, label the vertical axis as ln y, not y. (作图时纵轴标注 ln y，而不是 y。)
- If asked to give the answer in the form y = abᵗ, convert at the end with b = eᵏ. (若要求写成 y = abᵗ 形式，最后用 b = eᵏ 转换。)
- Interpret k with the correct sign and in the correct units. (解释 k 时注意正负号与单位。)
Finally, read the command word carefully. "Show that" requires a full algebraic derivation, "estimate" allows a graphical reading, and "comment on the validity" requires a sentence of criticism, not a calculation.
最后，要仔细阅读指令词。"证明"需要完整的代数推导，"估计"允许从图象读取，"评价有效性"需要写出一句批评性说明，而不是计算。
12. Summary and Key Formula Checklist | 小结与核心公式清单
The whole topic can be compressed into a short list of relationships. If you can move confidently between the differential equation, the e-form, the b-form and the log-linear form, you can handle almost any question the examiners set.
整个专题可以压缩成一份简短的关系清单。如果能在微分方程、e 形式、b 形式和对数线性形式之间自如转换，你就能应对考官设置的几乎所有题目。
| Relationship | Form |
| Differential equation | dy/dx = ky |
| Solution | y = aeᵏˣ |
| Change of base | y = abˣ, b = eᵏ, k = ln b |
| Linearised form | ln y = ln a + kx |
| Doubling or half-life | T = ln 2 ÷ |k| |
| Derivative and integral | dy/dx = ky, ∫ eᵏˣ dx = (1 ÷ k)eᵏˣ + C |
Practise by taking a real data set, plotting ln y against x, fitting a straight line and then converting the gradient and intercept back into a and k. That single exercise combines every skill this topic demands, and it is the fastest route to exam confidence.
练习方法：取一组真实数据，作 ln y 对 x 的图，拟合直线，再把斜率和截距还原为 a 和 k。这一练习综合了本专题要求的全部技能，是通往考试信心的最快途径。
Published by TutorHao | A-Level Mathematics Revision Series | aleveler.com
更多咨询请联系16621398022（同微信）
或截图后用「扫一扫 → 相册」选择本图