📚 Mathematical Modeling: Key Steps in Building Mathematical Models | 数学建模：建立数学模型的关键步骤
Mathematical modeling is the art of translating real-world problems into mathematical language, solving the resulting equations, and interpreting the solutions back in the context of the original problem. It is a core skill in A-Level mathematics and beyond, bridging abstract theory with practical application.
数学建模是将现实世界的问题转化为数学语言、求解所得方程并将答案还原到原始问题情境中的艺术。它是 A-Level 数学乃至更高层次学习的核心技能，连接了抽象理论与实际应用。
1. Problem Identification and Formulation | 问题识别与表述
The first and most crucial step is to clearly understand what question you are trying to answer. A vague problem such as “how fast does a disease spread?” must be refined into a precise question with measurable quantities, a time frame, and a defined population.
第一步也是最关键的一步，是明确你要回答的问题。像”疾病传播有多快”这样模糊的问题，必须被细化成一个带有可测量量、时间范围和明确人群的精准问题。
When identifying the problem, ask yourself the following questions:
在识别问题时，请问自己以下几个问题：
- What are the key quantities involved? | 涉及的关键量是什么？
- What is the time scale of interest? | 感兴趣的时间尺度是什么？
- What data or information is already available? | 已有哪些数据或信息？
- What level of accuracy is required? | 需要什么样的精度？
For example, if modelling the cooling of a cup of coffee, the problem might be stated as: “Given an initial temperature of 90 °C and room temperature of 20 °C, predict the temperature after 10 minutes.” This formulation gives you a clear target.
例如，若要对一杯咖啡的冷却过程建模，问题可以表述为：”已知初始温度为 90 °C，室温为 20 °C，预测 10 分钟后的温度。”这样的表述给出了清晰的目标。
2. Making Assumptions | 建立假设
No mathematical model can capture every detail of reality. Assumptions simplify the problem while preserving its essential features. They must be stated explicitly so that the model’s limitations are transparent.
没有任何数学模型能捕捉现实中的所有细节。假设可以在保留问题本质特征的同时简化问题。必须先明确陈述假设，使模型的局限性一目了然。
Common assumptions in A-Level modelling include:
A-Level 建模中常见的假设包括：
- Constant rates (e.g., a car travelling at a fixed speed). | 恒定速率（如汽车以固定速度行驶）。
- Uniform conditions (e.g., a well-mixed liquid). | 均匀条件（如充分混合的液体）。
- Negligible external influences (e.g., air resistance ignored). | 忽略外部影响（如忽略空气阻力）。
- Linear relationships where appropriate. | 在合适的地方使用线性关系。
To illustrate, Newton’s law of cooling assumes that the rate of temperature change is proportional to the temperature difference between the object and its surroundings. This ignores radiation effects and variations in room temperature, yet it produces remarkably accurate predictions for small temperature differences.
以牛顿冷却定律为例，它假设温度变化率与物体和周围环境之间的温差成正比。这忽略了辐射效应和室温的变化，但在温差较小时仍能给出相当准确的预测。
3. Defining Variables and Parameters | 定义变量与参数
Once assumptions are established, you must assign mathematical symbols to the quantities of interest. A variable changes within the system (such as time t or population N), while a parameter is a fixed constant that shapes the system (such as growth rate r).
假设确立后，你必须为所关注的量赋予数学符号。变量在系统内变化（如时间 t 或种群数量 N），而参数是塑造系统的固定常数（如增长率 r）。
A clear variable definition table is essential. For a population growth model, you might define:
清晰的变量定义表是必不可少的。对于种群增长模型，你可以定义：
| Symbol | 符号 | Meaning | 含义 | Type | 类型 |
| t | Time elapsed (hours) | 经过的时间（小时） | Independent variable | 自变量 |
| N(t) | Population at time t | t 时刻的种群数量 | Dependent variable | 因变量 |
| r | Per-capita growth rate | 人均增长率 | Parameter | 参数 |
Choosing the right variables is a skill that improves with practice. A good rule of thumb is to include only those quantities that are necessary and sufficient to answer the original question.
选择正确的变量是一项随着练习而提高的技能。一个好的经验法则是：只包含回答原始问题所必需且足够的量。
4. Constructing the Model | 构建模型
With variables defined, you now build equations that relate them. This is where mathematical knowledge comes into play: differential equations for rates of change, algebraic equations for static relationships, or statistical models for data-driven problems.
变量定义完成后，你需要建立联系它们的方程。这正是数学知识发挥作用的地方：变化率用微分方程，静态关系用代数方程，数据驱动问题用统计模型。
For a simple population model, the exponential growth equation is:
对于一个简单的种群模型，指数增长方程为：
dN/dt = rN
This differential equation states that the rate of change of the population is proportional to the current population. Its solution is an exponential function:
这个微分方程表明，种群的变化率与当前种群数量成正比。它的解是一个指数函数：
N(t) = N₀eʳᵗ
where N₀ is the initial population. Notice how each term in the equation corresponds to a real-world quantity, making the model both transparent and testable.
其中 N₀ 是初始种群数量。请注意，方程中的每一项都对应一个现实世界的量，这使得模型既透明又可检验。
5. Solving the Model | 求解模型
Solving a model means finding the mathematical expression that describes the behaviour of the system. Analytical solutions are exact — like the exponential solution above. Numerical solutions, such as Euler’s method or the Runge-Kutta method, provide approximations when analytical solutions are impossible.
求解模型意味着找出描述系统行为的数学表达式。解析解是精确的——如上面的指数解。数值解（如欧拉法或龙格-库塔法）在解析解无法求得时提供近似结果。
In A-Level mathematics, you are expected to be able to solve first-order differential equations by separation of variables. For the exponential model:
在 A-Level 数学中，你需要能够通过分离变量法求解一阶微分方程。对于指数模型：
∫ (1/N) dN = ∫ r dt
which gives ln|N| = rt + C, and hence N = N₀eʳᵗ after applying the initial condition N(0) = N₀.
从而得到 ln|N| = rt + C，并在应用初始条件 N(0) = N₀ 后得到 N = N₀eʳᵗ。
For more complex models, you may use numerical methods with a spreadsheet or a calculator. The key is to ensure your solution method matches the accuracy required by the original problem.
对于更复杂的模型，你可以使用电子表格或计算器进行数值求解。关键在于确保你的求解方法满足原始问题对精度的要求。
6. Model Validation | 模型验证
A model is only useful if it reflects reality. Validation involves comparing the model’s predictions with actual data or known results. If the model predicts a coffee temperature of 45 °C after 10 minutes but the measured value is 60 °C, the model needs revision.
模型只有在反映现实时才有用。验证涉及将模型的预测与实际数据或已知结果进行比较。如果模型预测咖啡 10 分钟后的温度为 45 °C，但实测值为 60 °C，则模型需要修正。
Key validation questions include:
关键的验证问题包括：
- Are the predictions within an acceptable error margin? | 预测是否在可接受的误差范围内？
- Does the model behave sensibly at extreme values? | 模型在极端值下是否表现合理？
- Does it pass intuitive sanity checks? | 它是否通过直觉合理性检查？
- Are the assumptions still valid over the range of interest? | 在所关注范围内假设是否仍然有效？
Often, the first attempt at validation reveals weaknesses. This is not a failure — it is part of the iterative nature of modelling, which we will explore in Section 9.
通常，第一次验证就会暴露模型的弱点。这不是失败——这是建模迭代性质的一部分，我们将在第 9 节中深入探讨。
7. Sensitivity Analysis | 灵敏度分析
Sensitivity analysis examines how changes in parameters affect the model’s output. This helps you understand which parameters are most influential and where uncertainty in data might cause the largest errors.
灵敏度分析考察参数变化如何影响模型的输出。这帮助你理解哪些参数影响最大，以及数据中的不确定性可能在何处引起最大的误差。
For the exponential growth model, the solution N(t) = N₀eʳᵗ depends on two parameters: N₀ and r. The sensitivity with respect to r is given by the partial derivative:
对于指数增长模型，解 N(t) = N₀eʳᵗ 依赖于两个参数：N₀ 和 r。对 r 的灵敏度由偏导数给出：
∂N/∂r = N₀t · eʳᵗ
This shows that sensitivity to the growth rate increases with time t — meaning that small errors in estimating r become large errors in long-term predictions.
这表明对增长率的灵敏度随时间 t 增大——意味着估计 r 时的微小误差会在长期预测中变成巨大误差。
In practice, sensitivity analysis guides where to invest effort in collecting better data and helps modellers communicate the reliability of their predictions.
在实践中，灵敏度分析指导我们应把精力投入在何处以收集更好的数据，并帮助建模者沟通其预测的可靠性。
8. Model Interpretation and Communication | 模型解释与交流
The solution of a mathematical model is meaningless until it is translated back into the language of the original problem. If the model predicts N(10) = 5,000 for a bacterial population, you must state: “After 10 hours, the predicted bacterial population is 5,000.”
模型的解在翻译回原始问题的语言之前毫无意义。如果模型预测 N(10) = 5,000 表示细菌种群，你必须说明：”10 小时后，预测的细菌种群数量为 5,000。”
Effective communication of model results involves:
有效的模型结果交流包括：
- Stating conclusions in plain language without jargon. | 用不含术语的平实语言陈述结论。
- Reporting confidence intervals or error margins when appropriate. | 在合适时报告置信区间或误差范围。
- Presenting graphs and tables that clearly illustrate key trends. | 提供清晰展示关键趋势的图表。
- Being honest about limitations and assumptions. | 诚实地说明局限性和假设。
This step is heavily assessed in A-Level examinations, particularly in the problem-solving and modelling questions, where clear written communication earns method marks even if calculations are imperfect.
这一步在 A-Level 考试中被重点考察，尤其是在问题求解和建模题中，即使计算不完美，清晰的书面表达也能获得方法分。
9. Iteration and Refinement | 迭代与改进
Mathematical modelling is rarely a one-shot process. After validation and sensitivity analysis, you typically return to earlier steps to refine assumptions, adjust parameters, or even redefine the problem itself.
数学建模很少是一蹴而就的过程。在验证和灵敏度分析之后，你通常会返回前面的步骤来改进假设、调整参数，甚至重新定义问题本身。
The modelling cycle can be represented as:
建模循环可以表示为：
Problem → Assumptions → Variables → Model → Solve → Validate → Refine → Repeat
问题 → 假设 → 变量 → 模型 → 求解 → 验证 → 改进 → 重复
For example, if the exponential growth model overpredicts population at large t, you might introduce a carrying capacity K to create the logistic model:
例如，如果指数增长模型在 t 较大时高估了种群数量，你可以引入环境容纳量 K 来建立逻辑斯蒂模型：
dN/dt = rN(1 − N/K)
This refined model limits growth as N approaches K, better reflecting real-world constraints. Each iteration improves the model until the desired accuracy is achieved.
这个改进后的模型在 N 接近 K 时限制增长，更好地反映了现实世界的约束。每一次迭代都会改进模型，直到达到所需的精度。
10. A Worked Example: Coffee Cooling | 实例分析：咖啡冷却
Let us apply the full modelling cycle to a classic problem. A cup of coffee at 90 °C is left in a room at 20 °C. After 5 minutes, it has cooled to 70 °C. Predict the temperature after 10 minutes.
让我们将完整的建模循环应用到一个经典问题上。一杯 90 °C 的咖啡被放置在 20 °C 的房间里。5 分钟后，它冷却到 70 °C。预测 10 分钟后的温度。
Step 1 — Problem: Predict the temperature T at time t = 10 minutes.
步骤 1 — 问题：预测 t = 10 分钟时的温度 T。
Step 2 — Assumptions: Newton’s law of cooling applies; room temperature is constant; the cup is well-mixed.
步骤 2 — 假设：牛顿冷却定律适用；室温恒定；咖啡充分混合。
Step 3 — Variables: T(t) = temperature at time t; Tₐ = 20 °C (room temperature); k = cooling constant.
步骤 3 — 变量：T(t) 为 t 时刻的温度；Tₐ = 20 °C（室温）；k 为冷却常数。
Step 4 — Model: The differential equation is:
步骤 4 — 模型：微分方程为：
dT/dt = −k(T − Tₐ)
Step 5 — Solve: Separation of variables gives the general solution:
步骤 5 — 求解：分离变量得到通解：
T(t) = Tₐ + (T₀ − Tₐ)e⁻ᵏᵗ
Using T₀ = 90 °C and Tₐ = 20 °C: T(t) = 20 + 70e⁻ᵏᵗ. From T(5) = 70, we solve 70 = 20 + 70e⁻⁵ᵏ, which gives e⁻⁵ᵏ = 5/7, so k = −(1/5)ln(5/7) ≈ 0.0673 min⁻¹. Therefore:
使用 T₀ = 90 °C 和 Tₐ = 20 °C：T(t) = 20 + 70e⁻ᵏᵗ。由 T(5) = 70，解 70 = 20 + 70e⁻⁵ᵏ，得 e⁻⁵ᵏ = 5/7，所以 k = −(1/5)ln(5/7) ≈ 0.0673 分钟⁻¹。因此：
T(10) = 20 + 70e^(−0.0673 × 10) ≈ 20 + 70 × 0.511 ≈ 55.8 °C
Step 6 — Validate: If you actually measure the coffee, you might find 57 °C — the difference may arise from assumptions like neglecting the cooling of the cup itself. Sensitivity analysis would tell you that the prediction depends strongly on the measurement at t = 5, so taking multiple data points would improve reliability.
步骤 6 — 验证：如果你实际测量咖啡温度，可能得到 57 °C——差异可能源于忽略杯子本身冷却等假设。灵敏度分析会告诉你，预测强烈依赖于 t = 5 时的测量，因此取多个数据点会提高可靠性。
11. Common Pitfalls in Mathematical Modelling | 数学建模中的常见误区
Students frequently make avoidable mistakes when building models. Being aware of these pitfalls will improve your modelling skills significantly.
学生在建模时经常会犯一些可以避免的错误。了解这些误区将显著提高你的建模技能。
- Overcomplicating: Adding too many variables makes the model unmanageable without improving accuracy. | 过度复杂化：加入过多变量会使模型难以管理，且不提高精度。
- Unchecked assumptions: Failing to test whether assumptions hold in the situation being modelled. | 未经检验的假设：未能检验假设在建模情境中是否成立。
- Unit errors: Mixing units (e.g., hours vs. minutes) leads to incorrect results. | 单位错误：混用单位（如小时和分钟）会导致错误结果。
- Circular reasoning: Using the answer to prove the model without independent validation. | 循环论证：用答案来证明模型，而没有独立验证。
- Ignoring errors: Not considering measurement or numerical approximation errors in the final answer. | 忽略误差：没有在最终答案中考虑测量或数值近似误差。
Always revisit the original problem statement before finalising a model. The best model is the simplest one that answers the question with acceptable accuracy.
在最终确定模型之前，一定要重新审视原始问题陈述。最好的模型是能以可接受精度回答问题的最简单模型。
12. Conclusion | 总结
Mathematical modelling is a powerful, iterative process that transforms real-world questions into rigorous mathematics. The key steps — problem identification, assumptions, variable definition, model construction, solving, validation, sensitivity analysis, and iteration — form a cycle that can be applied to virtually any quantitative problem.
数学建模是一个强大的、迭代的过程，它将现实世界的问题转化为严谨的数学。关键步骤——问题识别、假设、变量定义、模型构建、求解、验证、灵敏度分析和迭代——形成了一个几乎适用于任何定量问题的循环。
For A-Level students, mastering these steps not only earns marks in examinations but also develops transferable skills for university study and professional careers in science, engineering, economics, and data science. Start with simple models, practise the full cycle, and always ask: “Does my model make sense?”
对于 A-Level 学生来说，掌握这些步骤不仅在考试中获得分数，还能培养可在大学学习以及科学、工程、经济学和数据科学等职业生涯中迁移的技能。从简单模型开始，练习完整的循环，并始终问自己：”我的模型合理吗？”
Published by TutorHao | Mathematics Revision Series | aleveler.com
更多咨询请联系16621398022（同微信）
屏轩国际教育cambridge primary/secondary checkpoint, cat4, ukiset,ukcat,igcse,alevel,PAT,STEP,MAT, ibdp,ap,ssat,sat,sat2课程辅导，国外大学本科硕士研究生博士课程论文辅导Cancel reply