7.3 Recurrence Relations and General Term Determination

Key Takeaways

  • For additive recurrences a_{n+1} - a_n = f(n), the difference accumulation method gives a_n = a_1 + Sum_{k=1}^{n-1} f(k).
  • For multiplicative recurrences a_{n+1}/a_n = f(n), the product accumulation method gives a_n = a_1 * Prod_{k=1}^{n-1} f(k).
  • First-order linear recurrences a_{n+1} = p a_n + q (p != 1) are solved by undetermined coefficient transformation a_{n+1} - lambda = p(a_n - lambda) with lambda = q / (1-p).
  • When given S_n relations, apply a_n = S_n - S_{n-1} for n >= 2, always verifying whether the initial term a_1 = S_1 conforms to the general pattern.
  • Recurrences involving variable additions like a_{n+1} = p a_n + k r^n can be linearized by dividing by r^{n+1} to create an auxiliary sequence.
Last updated: July 2026

7.3 Recurrence Relations and General Term Determination

Determining the general term $a_n$ from a recurrence relation is a central topic in Gaokao Mathematics. This section provides a comprehensive classification of standard recurrence structures and their exact resolution algorithms.

1. Difference Accumulation Method (累加法)

Applicable Type

Recurrences of the form:

an+1an=f(n)a_{n+1} - a_n = f(n)

where $f(n)$ is a solvable explicit function of $n$ (such as a polynomial, exponential, or rational expression whose sum $\sum f(k)$ can be computed).

Algorithm

Write out the equations for successive indices up to $n-1$:

a_2 - a_1 &= f(1) \\ a_3 - a_2 &= f(2) \\ &\vdots \\ a_n - a_{n-1} &= f(n-1) \end{aligned}$$ Summing all $n-1$ equations produces a telescoping sum on the left: $$(a_n - a_{n-1}) + \dots + (a_2 - a_1) = \sum_{k=1}^{n-1} f(k) \implies a_n = a_1 + \sum_{k=1}^{n-1} f(k) \quad (n \ge 2)$$ > **Mandatory Check**: Always verify whether the derived formula holds for $n=1$. ## 2. Product Accumulation Method (累乘法) ### Applicable Type Recurrences of the form: $$\frac{a_{n+1}}{a_n} = f(n) \quad (a_n \neq 0, f(n) \neq 0)$$ where the product $\prod_{k=1}^{n-1} f(k)$ simplifies via cancellation or known multiplicative formulas. ### Algorithm Write out successive ratios for $k=1, 2, \dots, n-1$: $$\frac{a_2}{a_1} = f(1), \quad \frac{a_3}{a_2} = f(2), \quad \dots, \quad \frac{a_n}{a_{n-1}} = f(n-1)$$ Multiplying all $n-1$ ratios together yields: $$\frac{a_n}{a_1} = \prod_{k=1}^{n-1} f(k) \implies a_n = a_1 \cdot \prod_{k=1}^{n-1} f(k) \quad (n \ge 2)$$

3. First-Order Linear Recurrence $a_{n+1} = p a_n + q$ (待定系数构造法)

When $p = 1$, the relation reduces to $a_{n+1} - a_n = q$ (arithmetic sequence). When $q = 0$, it reduces to $a_{n+1} = p a_n$ (geometric sequence). When $p \neq 1$ and $q \neq 0$, we construct an auxiliary geometric sequence.

Undetermined Coefficient Transformation

Assume the recurrence can be written in the form:

an+1λ=p(anλ)a_{n+1} - \lambda = p(a_n - \lambda)

Expanding this equation: $a_{n+1} = p a_n + \lambda(1 - p)$. Comparing coefficients with $a_{n+1} = p a_n + q$ yields:

λ(1p)=q    λ=q1p\lambda(1 - p) = q \implies \lambda = \frac{q}{1 - p}

Define the auxiliary sequence $b_n = a_n - \lambda$. Then $b_1 = a_1 - \lambda$, and $b_{n+1} = p b_n$. Thus, $(b_n)$ is a geometric sequence with initial term $b_1$ and common ratio $p$:

bn=b1pn1=(a1λ)pn1b_n = b_1 \cdot p^{n-1} = (a_1 - \lambda) p^{n-1}

Substituting back $a_n = b_n + \lambda$ gives the general term:

an=(a1λ)pn1+λ=(a1q1p)pn1+q1pa_n = (a_1 - \lambda) p^{n-1} + \lambda = \left(a_1 - \frac{q}{1-p}\right) p^{n-1} + \frac{q}{1-p}

4. Higher Recurrence Structures: $a_{n+1} = p a_n + f(n)$

Function Type $f(n)$Auxiliary Transformation StrategyResulting Auxiliary Sequence
$f(n) = k r^n$ ($r \neq p$)Divide both sides by $r^{n+1}$: $\frac{a_{n+1}}{r^{n+1}} = \frac{p}{r} \frac{a_n}{r^n} + \frac{k}{r}$Let $b_n = \frac{a_n}{r^n}$, reducing to $b_{n+1} = p' b_n + q'$
$f(n) = k p^n$ ($r = p$)Divide both sides by $p^{n+1}$: $\frac{a_{n+1}}{p^{n+1}} = \frac{a_n}{p^n} + \frac{k}{p}$Let $b_n = \frac{a_n}{p^n}$, giving AP $b_{n+1} - b_n = \frac{k}{p}$
$f(n) = A n + B$Construct $a_{n+1} + X(n+1) + Y = p(a_n + Xn + Y)$Match coefficients to solve for $X$ and $Y$

5. Sequence Determination via $S_n$ and $a_n$ Relations

In many Gaokao problems, the defining condition involves both $S_n$ and $a_n$.

Fundamental Bridge

an={S1,n=1SnSn1,n2a_n = \begin{cases} S_1, & n = 1 \\ S_n - S_{n-1}, & n \ge 2 \end{cases}

Core Resolution Pathways:

  1. Eliminate $S_n$ to find $a_n$ relation: Write the relation for index $n$: $f(S_n, a_n) = 0$. Write the relation for index $n-1$ ($n \ge 2$): $f(S_{n-1}, a_{n-1}) = 0$. Subtract the two equations and substitute $S_n - S_{n-1} = a_n$ to obtain a pure recurrence relation in $a_n$.
  2. Eliminate $a_n$ to find $S_n$ relation: Substitute $a_n = S_n - S_{n-1}$ ($n \ge 2$) directly into the given equation to obtain a recurrence relation in $S_n$, find $S_n$, and then recover $a_n = S_n - S_{n-1}$.
Loading diagram...
Recurrence Relation Solution Taxonomy

Worked Gaokao Exam Examples

Example 1 (Undetermined Coefficients for $a_{n+1} = p a_n + q$)

Problem: A sequence $(a_n)$ satisfies $a_1 = 2$ and $a_{n+1} = 3 a_n - 2$ for all $n \ge 1$.

  1. Show that $(a_n - 1)$ is a geometric sequence and state its common ratio.
  2. Find the general term formula $a_n$.

Solution:

  1. We construct $\lambda$ such that $a_{n+1} - \lambda = 3(a_n - \lambda)$. Expanding gives $a_{n+1} = 3a_n - 2\lambda$. Comparing with $a_{n+1} = 3a_n - 2$: 2λ=2    λ=1-2\lambda = -2 \implies \lambda = 1 Thus, $a_{n+1} - 1 = 3(a_n - 1)$ for all $n \ge 1$. Calculate the initial term of $(a_n - 1)$: b1=a11=21=1b_1 = a_1 - 1 = 2 - 1 = 1 Since $\frac{a_{n+1} - 1}{a_n - 1} = 3$ is a constant ratio, $(a_n - 1)$ is a geometric sequence with initial term $1$ and common ratio $q = 3$.

  2. The general term of the auxiliary sequence is: an1=13n1=3n1a_n - 1 = 1 \cdot 3^{n-1} = 3^{n-1} Adding 1 gives the general term $a_n$: an=3n1+1a_n = 3^{n-1} + 1


Example 2 (General Term from $S_n$ Recurrence Relation)

Problem: Let $(a_n)$ be a sequence with partial sum $S_n$ satisfying $S_n = 2 a_n - 2$ for all $n \ge 1$.

  1. Find $a_1$ and $a_2$.
  2. Determine the general term formula $a_n$.

Solution:

  1. For $n = 1$, $S_1 = a_1$, so: a1=2a12    a1=2a_1 = 2a_1 - 2 \implies a_1 = 2 For $n = 2$, $S_2 = a_1 + a_2 = 2 + a_2$. Using $S_2 = 2a_2 - 2$: 2+a2=2a22    a2=42 + a_2 = 2a_2 - 2 \implies a_2 = 4

  2. For $n \ge 2$, we have: Sn=2an2S_n = 2a_n - 2 Sn1=2an12S_{n-1} = 2a_{n-1} - 2 Subtracting the second equation from the first: SnSn1=2an2an1S_n - S_{n-1} = 2a_n - 2a_{n-1} Since $S_n - S_{n-1} = a_n$ for $n \ge 2$: an=2an2an1    an=2an1(n2)a_n = 2a_n - 2a_{n-1} \implies a_n = 2a_{n-1} \quad (n \ge 2) Thus, $(a_n)$ is a geometric sequence with first term $a_1 = 2$ and common ratio $q = 2$. The general term formula is: an=22n1=2na_n = 2 \cdot 2^{n-1} = 2^n

Test Your Knowledge

Given a_1 = 2 and a_{n+1} = 3a_n - 2, what is the general term a_n?

A
B
C
D
Test Your Knowledge

A sequence satisfies a_1 = 1 and a_{n+1} - a_n = 2n. What is the value of a_10?

A
B
C
D
Test Your Knowledge

For a sequence (a_n) with partial sum S_n, if S_n = 3^n - 1, what is a_n for n >= 1?

A
B
C
D