Proof. The function F, viewed as a continuous function with respect to x1,x2,⋯,x11, has a maximum value on the bounded closed set
Ω={(x1,x2,⋯,x11)∈(R≥0)11∣x1+x2+⋯+x11=1}.
Assume that F has a maximum value at (a1,a2,⋯,a11)∈Ω. It is clear that F>0 at this point.
(1) a3=a5=a7=a9=a11=0.
If a3>0, let a3′=0, a4′=a3+a4, and let ai′=ai for other i. This maintains the sum. We prove that F(a1,a2,⋯,a11)<F(a1′,a2′,⋯,a11′), which is equivalent to
(a2+a3+a4)(a3+a4)(a4+a5+a6)<(a2′+a3′+a4′)(a3′+a4′)(a4′+a5′+a6′),
and since a3+a4=a3′+a4′ and a4<a4′, this inequality holds. This contradicts the fact that F has a maximum value at (a1,a2,⋯,a11). Therefore, a3=0. Similarly, we can prove that a5=a7=a9=a11=0.
Now, let x3=x5=x7=x9=x11=0 in the expression for F. We will now view F as a function of x1,x2,x4,⋯,x10,
F=(x1+x2)(x2+x4)x4(x4+x6)x6(x6+x8)x8(x8+x10)x10(x10+x1)x1.
Then, F has a maximum value at (a1,a2,a4,a6,a8,a10). It is clear that a1,a4,a6,a8,a10>0.
(2) a2=0.
By contradiction, suppose a2>0. If a1≤a4, let a1′=a1+a2, a2′=0, and let ai′=ai for other i. This maintains the sum. We prove that F(a1,a2,a4,⋯,a10)<F(a1′,a2′,a4′,⋯,a10′).
⇔a1(a10+a1)(a1+a2)(a2+a4)<a1′(a10′+a1′)(a1′+a2′)(a2′+a4′).
Since a1<a1′ and a10+a1<a10′+a1′, we only need to prove:
a1(a1+a2)(a2+a4)≤a1′(a1′+a2′)(a2′+a4′)=(a1+a2)(a1+a2)a4,
⇔a1(a2+a4)≤a4(a1+a2),⇔a1a2≤a4a2,
which holds. If a1≥a4, let a4′=a4+a2, a2′=0, and for other i, let ai′=ai. We also have:
F(a1,a2,a4,⋯,a10)<F(a1′,a2′,a4′,⋯,a10′).
This contradicts the fact that F reaches its maximum value at (a1,a2,a4,a6,a8,a10). Therefore, a2=0.
Now let x2=0 in the analytic expression of F, and regard F as a function of x1,x4,x6,x8,x10,
F=x42(x4+x6)x6(x6+x8)x8(x8+x10)x10(x10+x1)x12.
Then F reaches its maximum value at (a1,a4,a6,a8,a10).
(3) a1=a4,a6=a10.
Let a1′=a4′=21(a1+a4), a6′=a10′=21(a6+a10), a8′=a8, keeping the sum unchanged. By the inequality of arithmetic and geometric means,
a 1 2 a 4 2 a’ 1 2 a’ 4 2, (a 4 + a 6)(a 10 + a 1) (a’ 4 + a’ 6)(a’ 10 + a’ 1),
a6a10≤a6′a10′,(a6+a8)(a8+a10)≤(a6′+a8′)(a8′+a10′).
Also, a8=a8′. Multiplying these equations gives:
F(a1,a4,…,a10)≤F(a1′,a4′,…,a10′).
Since F reaches its maximum value at (a1,a4,…,a10), the equality holds in the above inequality. By the conditions for equality in the inequality of arithmetic and geometric means, we have a1=a4,a6=a10.
In the analytic expression of F, replace x4 with x1, and x10 with x6. Now regard F as a function of x1,x6,x8,
F=x14x62x8(x1+x6)2(x6+x8)2,
and 2x1+2x6+x8=1.
(4) Next, we will solve the system of equations (this system of equations is obtained by matching coefficients in the inequality of arithmetic and geometric means, which will be used in (5)):
u2+u+v1=v1+u+v1+v+11=1+v+12,u,v>0.
We will prove that there is a unique positive real solution u,v, and that v<1.
From the first equation, we get u2=v1+v+11, hence u=2v+12v(v+1). From the second equation, we get v1+u+v1=1+v+11, substituting u, and rearranging it as an equation about v,
4v3+5v2−4v−4=0.
Let f(t)=4t3+5t2−4t−4, then f′(t)=12t2+10t−4, which is negative at first and positive later on [0,+∞), so f(t) first decreases and then increases on [0,+∞). Since f(0)=−4<0, f(1)=1>0, therefore, f has a unique solution v∈(0,1) on [0,+∞), and u is also uniquely determined.
(5) Let u,v>0 be the solution satisfying the system of equations in (4), and let u2+u+v1=k.
Using the inequality of arithmetic and geometric means, we have:
F(x1,x6,x8)=(ux1)4(vx6)2x8(u+vx1+x6)2(v+1x6+x8)2u4v2(u+v)2(v+1)2≤1111u4v2(u+v)2(v+1)2(u4x1+v2x6+x8+u+v2(x1+x6)+v+12(x6+x8))11=1111u4v2(u+v)2(v+1)2((u4+u+v2)x1+(v2+u+v2+v+12)x6+(1+v+12)x8)11=1111u4v2(u+v)2(v+1)2(2kx1+2kx6+kx8)11=1111u4v2(u+v)2(v+1)2k11.
The equality above holds if and only if x1:x6:x8=u:v:1, that is, x8=2u+2v+11,
x6=2u+2v+1v, x1=2u+2v+1u. Therefore, F attains its maximum value if and only if x2=x3=
x5=x7=x9=x11=0, x1=x4=2u+2v+1u, x6=x10=2u+2v+1v, x8=2u+2v+11. At this point,
x6=vx8<x8. □