返回题库

AIME 2007 II · 第 14 题

AIME 2007 II — Problem 14

专题
Contest Math
难度
L4
来源
AIME

题目详情

Problem

Let f(x)f(x) be a polynomial with real coefficients such that f(0)=1,f(0) = 1, f(2)+f(3)=125,f(2)+f(3)=125, and for all xx, f(x)f(2x2)=f(2x3+x).f(x)f(2x^{2})=f(2x^{3}+x). Find f(5).f(5).

Official Solution (MAA)

If the leading term of f(x)f(x) is axmax^m, then the leading term of f(x)f(2x2)=axma(2x2)m=2ma2x3mf(x)f(2x^2) = ax^m \cdot a(2x^2)^m = 2^ma^2x^{3m}, and the leading term of f(2x3+x)=2max3mf(2x^3 + x) = 2^max^{3m}. Hence 2ma2=2ma2^ma^2 = 2^ma, and a=1a = 1. Because f(0)=1f(0) = 1, the product of all the roots of f(x)f(x) is ±1\pm 1. If f(λ)=0f(\lambda) = 0, then f(2λ3+λ)=0f(2\lambda^3 + \lambda) = 0. Assume that there exists a root λ\lambda with λ1|\lambda| \neq 1. Then there must be such a root λ1\lambda_1 with λ1>1|\lambda_1| > 1. Then

2λ3+λ2λ3λ>2λλ=λ.|2\lambda^3 + \lambda| \ge 2|\lambda|^3 - |\lambda| > 2|\lambda| - |\lambda| = |\lambda|. But then f(x)f(x) would have infinitely many roots, given by λk+1=2λk3+λk\lambda_{k+1} = 2\lambda_k^3 + \lambda_k, for k1k \ge 1. Therefore λ=1|\lambda| = 1 for all of the roots of the polynomial. Thus λλ=1\lambda\overline{\lambda}=1, and (2λ3+λ)(2λ3+λ)=1(2\lambda^3 + \lambda)\overline{(2\lambda^3 + \lambda)} = 1. Solving these equations simultaneously for λ=a+bi\lambda = a + bi yields a=0a = 0, b2=1b^2 = 1, and so λ2=1\lambda^2 = -1. Because the polynomial has real coefficients, the polynomial must have the form f(x)=(1+x2)nf(x) = (1 + x^2)^n for some integer n1n \ge 1. The condition f(2)+f(3)=125f(2) + f(3) = 125 implies n=2n = 2, giving f(5)=676f(5) = \boxed{676}.

解析

Solution 1

Let rr be a root of f(x)f(x). Then we have f(r)f(2r2)=f(2r3+r)f(r)f(2r^2)=f(2r^3+r); since rr is a root, we have f(r)=0f(r)=0; therefore 2r3+r2r^3+r is also a root. Thus, if rr is real and non-zero, 2r3+r>r|2r^3+r|>r, so f(x)f(x) has infinitely many roots. Since f(x)f(x) is a polynomial (thus of finite degree) and f(0)f(0) is nonzero, f(x)f(x) has no real roots.

Note that f(x)f(x) is not constant. We then find two complex roots: r=±ir = \pm i. We find that f(i)f(2)=f(i)f(i)f(-2) = f(-i), and that f(i)f(2)=f(i)f(-i)f(-2) = f(i). This means that f(i)f(i)f(2)2=f(i)f(i)f(i)f(i)(f(2)21)=0f(i)f(-i)f(-2)^2 = f(i)f(-i) \Longrightarrow f(i)f(-i)(f(-2)^2 - 1) = 0. Thus, ±i\pm i are roots of the polynomial, and so (xi)(x+i)=x2+1(x - i)(x + i) = x^2 + 1 will be a factor of the polynomial. (Note: This requires the assumption that f(2)1f(-2)\neq1. Clearly, f(2)1f(-2)\neq-1, because that would imply the existence of a real root.)

The polynomial is thus in the form of f(x)=(x2+1)g(x)f(x) = (x^2 + 1)g(x). Substituting into the given expression, we have

(x2+1)g(x)(4x4+1)g(2x2)=((2x3+x)2+1)g(2x3+x)(x^2+1)g(x)(4x^4+1)g(2x^2)=((2x^3+x)^2+1)g(2x^3+x) (4x6+4x4+x2+1)g(x)g(2x2)=(4x6+4x4+x2+1)g(2x3+x)(4x^6+4x^4+x^2+1)g(x)g(2x^2)=(4x^6+4x^4+x^2+1)g(2x^3+x) Thus either 4x6+4x4+x2+1=(4x4+1)(x2+1)4x^6+4x^4+x^2+1=(4x^4+1)(x^2+1) is 0 for any xx, or g(x)g(x) satisfies the same constraints as f(x)f(x). Continuing, by infinite descent, f(x)=(x2+1)nf(x) = (x^2 + 1)^n for some nn.

Since f(2)+f(3)=125=5n+10nf(2)+f(3)=125=5^n+10^n for some nn, we have n=2n=2; so f(5)=676f(5) = \boxed{676}.

Comment: The answer is clearly correct, but the proof has a gap, i.e. there is no reason that f(2)1f(-2)\neq1. Since f(x)f(x) has no real roots, the degree must be even. Consider g(x)=f(x)/f(x)g(x)= f(x)/f(-x). Then since ff is non-zero, g(x)=g(2x3+x)g(x)=g(2x^3+x). Now the function 2x3+x2x^3+x applied repeatedly from some real starting value of x becomes arbitrarily large, and the limit of g(x)g(x) as x|x| approaches infinity is 1, so g(x)g(x)=1 for all x, or f(x)=f(x)f(x)=f(-x). Then f(x)=h(x2+1)f(x)=h(x^2+1) for some polynomial h(x)h(x), and h(x2+1)h(4x4+1)=h(4x6+4x4+x2+1)=h((x2+1)(4x4+1))h(x^2+1)h(4x^4+1)=h(4x^6+4x^4+x^2+1) = h((x^2+1)(4x^4+1)). Now suppose h has degree m. It is clearly monic. Assume that the next highest non-zero coefficient in h is k. Then, subtracting ((x2+1)(4x4+1))m((x^2+1)(4x^4+1))^m from both sides of the equation yields a polynomial equality with degree 4m+2k4m+2k on the left and degree 6k6k on the right, a contradiction. So h(x)=xmh(x)=x^m, and f(x)=(1+x2)mf(x)=(1+x^2)^m.

Solution 2

Let rr be a root of f(x).f(x). This means that f(r)f(2r2)=f(2r3+r).f(r)f(2r^2)=f(2r^3+r). In other words, 2r3+r2r^3+r is a root of f(x)f(x) too. Since f(x)f(x) can't have infinitely many roots,

Q(x)=P(P(P(P(r))))Q(x)=P(P(\dotsb P(P(r)) \dotsb)) is cyclic, where P(x)=2x3+x.P(x)=2x^3+x. Now, we will do casework.

Case 1: degf1\deg f\geq1

Subcase 1: r>1|r|>1

This means that

2r3+r2r3r=r(2r21)>r(2121)=r.|2r^3+r|\geq|2r^3|-|r|=|r|(2|r|^2-1)>|r|(2\cdot1^2-1)=|r|. It follows that 2r3+r>r|2r^3+r|>|r| for all r.r. This implies that Q(r)Q(r) can't be cyclic. Thus, it is impossible for r>1|r|>1 to be true.

Subcase 2: r<1|r|<1

This means that 2r3+r2r3r=r(2r21)<r.|2r^3+r|\geq2|r^3|-|r|=|r|(|2r^2|-1)<|r|. It follows that 2r3+r<r|2r^3+r|<|r| for all r.r. This implies that Q(r)Q(r) can't be cyclic. Thus, it is impossible for r>1|r|>1 to be true.

Subcase 3: r=1.|r|=1.

Since r|r| is not greater than or less than 1, r=1.|r|=1. This means that all the roots of the polynomial have a magnitude of 1.1. More specifically, 2r3+r|2r^3+r| has a magnitude of one. Since this would mean an equality condition from the triangle inequality, 2r32r^3 and rr are collinear with the origin in the complex plane. In other words, 2r3r=±ccr=2r32r2=cr=±±c2,\frac{2r^3}{r}=\pm c\Leftrightarrow cr=2r^3\Leftrightarrow 2r^2=c\Leftrightarrow r=\pm\sqrt{\pm\frac{c}{2}}, for some real constant c.c. Now, from r=1,|r|=1, we find that ±±c2=1±c2=1±c2=1c=±2.\left|\pm\sqrt{\pm\frac{c}{2}}\right|=1\Leftrightarrow \sqrt{\pm\frac{c}{2}}=1\Leftrightarrow \pm\frac{c}{2}=1\Leftrightarrow c=\pm2. Putting this back into the equation, we find that r=1,1,i,i.r=1,-1,i,-i. Now, this means that 2r3+r=3,3,i,i.2r^3+r=3,-3,i,-i. 33 and 3-3 obviously doesn't have a magnitude of 1.1. Thus, i,ii,-i are the only possible roots of the polynomial. Since roots come in conjugate pairs, f(x)=[(xi)(x+i)]n=(x2+1)n,f(x)=[(x-i)(x+i)]^n=(x^2+1)^n, works for all constants n0.n\neq0.

Case 2: degf=0.\deg f=0.

This means that f(x)=c,f(x)=c, for some constant c.c. In other words, c2=c.c^2=c. We can easily find that this means that c=0,1.c=0,1. Combining all the cases, we conclude that f(x)=(x2+1)n,0,1f(x)=(x^2+1)^n,0,1 are the only polynomials that satisfy this equation. Now, we can test! f(x)=0,1f(x)=0,1 obviously don't satisfy f(2)+f(3)=125.f(2)+f(3)=125. Thus, f(x)=(x2+1)n.f(x)=(x^2+1)^n. Substituting, we find that 5n+10n=125n=2.5^n+10^n=125\Leftrightarrow n=2. We conclude that f(5)=(52+1)2=262=676.f(5)=(5^2+1)^2=26^2=\boxed{676}.

~ pinkpig

Video Solution

2007 AIME II #14

MathProblemSolvingSkills.com